Experimental Neurotoxicology Methods (Neuromethods, 172) 1071616366, 9781071616369

Citation preview

Neuromethods 172

Jordi Llorens Marta Barenys Editors

Experimental Neurotoxicology Methods

NEUROMETHODS

Series Editor Wolfgang Walz University of Saskatchewan, Saskatoon, SK, Canada

For further volumes: http://www.springer.com/series/7657

Neuromethods publishes cutting-edge methods and protocols in all areas of neuroscience as well as translational neurological and mental research. Each volume in the series offers tested laboratory protocols, step-by-step methods for reproducible lab experiments and addresses methodological controversies and pitfalls in order to aid neuroscientists in experimentation. Neuromethods focuses on traditional and emerging topics with wide-ranging implications to brain function, such as electrophysiology, neuroimaging, behavioral analysis, genomics, neurodegeneration, translational research and clinical trials. Neuromethods provides investigators and trainees with highly useful compendiums of key strategies and approaches for successful research in animal and human brain function including translational “bench to bedside” approaches to mental and neurological diseases.

Experimental Neurotoxicology Methods Edited by

Jordi Llorens Departament de Ciències Fisiològiques, Institut de Neurociènces, Universitat de Barcelona, Hospitalet de Llobregat, Catalunya, Spain; Institut d’Investigació Biomèdica de Bellvitge, IDIBELL, Hospitalet de Llobregat, Catalunya, Spain

Marta Barenys Departament de Farmacologia, Toxicologia i Química Terapèutica, Universitat de Barcelona, Barcelona, Catalunya, Spain

Editors Jordi Llorens Departament de Cie`ncies Fisiolo`giques, Institut de Neurocie`nces Universitat de Barcelona, Hospitalet de Llobregat Catalunya, Spain

Marta Barenys Departament de Farmacologia, Toxicologia i Quı´mica Terape`utica Universitat de Barcelona Barcelona, Catalunya, Spain

Institut d’Investigacio´ Biome`dica de Bellvitge IDIBELL, Hospitalet de Llobregat Catalunya, Spain

ISSN 0893-2336 ISSN 1940-6045 (electronic) Neuromethods ISBN 978-1-0716-1636-9 ISBN 978-1-0716-1637-6 (eBook) https://doi.org/10.1007/978-1-0716-1637-6 © Springer Science+Business Media, LLC, part of Springer Nature 2021, Corrected Publication 2023 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Cover Caption: This image is courtesy of Marta Barenys. This Humana imprint is published by the registered company Springer Science+Business Media, LLC, part of Springer Nature. The registered company address is: 1 New York Plaza, New York, NY 10004, U.S.A.

Preface to the Series Experimental life sciences have two basic foundations: concepts and tools. The Neuromethods series focuses on the tools and techniques unique to the investigation of the nervous system and excitable cells. It will not, however, short-change the concept side of things as care has been taken to integrate these tools within the context of the concepts and questions under investigation. In this way, the series is unique in that it not only collects protocols but also includes theoretical background information and critiques which led to the methods and their development. Thus, it gives the reader a better understanding of the origin of the techniques and their potential future development. The Neuromethods publishing program strikes a balance between recent and exciting developments like those concerning new animal models of disease, imaging, in vivo methods, and more established techniques, including, for example, immunocytochemistry and electrophysiological technologies. New trainees in neurosciences still need a sound footing in these older methods in order to apply a critical approach to their results. Under the guidance of its founders, Alan Boulton and Glen Baker, the Neuromethods series has been a success since its first volume published through Humana Press in 1985. The series continues to flourish through many changes over the years. It is now published under the umbrella of Springer Protocols. While methods involving brain research have changed a lot since the series started, the publishing environment and technology have changed even more radically. Neuromethods has the distinct layout and style of the Springer Protocols program, designed specifically for readability and ease of reference in a laboratory setting. The careful application of methods is potentially the most important step in the process of scientific inquiry. In the past, new methodologies led the way in developing new disciplines in the biological and medical sciences. For example, Physiology emerged out of Anatomy in the nineteenth century by harnessing new methods based on the newly discovered phenomenon of electricity. Nowadays, the relationships between disciplines and methods are more complex. Methods are now widely shared between disciplines and research areas. New developments in electronic publishing make it possible for scientists that encounter new methods to quickly find sources of information electronically. The design of individual volumes and chapters in this series takes this new access technology into account. Springer Protocols makes it possible to download single protocols separately. In addition, Springer makes its print-on-demand technology available globally. A print copy can therefore be acquired quickly and for a competitive price anywhere in the world.

v

Preface Neurotoxicology is devoted to study the adverse effects of chemical and physical agents on the structure or function of the nervous system. At the halfway between toxicology and neuroscience, it has developed its own identity, approaches, and methods. These methods are sometimes adaptations of methods first developed in other areas of neuroscience. In other cases, however, they are original methods developed to respond to specific needs of neurotoxicology, and these have remained within this domain or gained wider use. This volume includes a collection of detailed protocols currently used in neurotoxicology. It encompasses some well-established methods that constitute gold-standards to approach key questions frequently addressed in neurotoxicological research. Other chapters describe protocols for new approaches that appear as important advancements in the field called to have an impact on its development. The number of chapters that could be included in this volume is enormous; therefore, a strategic selection has been made. This selection provides an overall vision of the field, illustrating the variety of available approaches, from the molecular level (Chapts .3, 12, 13, 14, and 20) to the organism level (Chaps. 5–11, 15–18), by way of methods at the cellular (Chaps. 1–4, 13, 19–22) and tissue/organ level (chapters 2, 4–7, 10, 18, and 19). For the sake of completeness of the view, both in vitro and in vivo methods have been included. Nevertheless, the selection has been biased favoring in vivo methods because in vitro methods in neurotoxicology have been addressed in other recent books. Therefore, in vitro methods are included to provide the intended broad coverage, but they are represented less than their present weight in the field. Among organisms, we have covered methods based on a broad range of species, including the most common laboratory animals, rodents, and alternative organism models, but also human cells, thus circumventing the need of interspecies extrapolation. Globally, we aim to provide a tool useful to many laboratories, either new or established in the field, which may need to implement new protocols to address neurotoxicological questions. The 22 chapters included in the volume have been written by researchers that are recognized internationally for their expertise in the field and who are describing methods of current use in their laboratories. They have been peer-reviewed and revised according to the reviewer’s criticisms. Therefore, each chapter provides an authoritative view and in-depth knowledge on the specific method being covered. Most chapters include detailed technical information for thorough reproducibility of the methods, as well as ample discussion of their applicability domain, and their advantages and limitations, with the aim of settling knowledge of the field and make it progress. This will make them useful to both young and experienced researchers, as an authoritative guidance for implementing any of these methods to the reader’s laboratory or for a critical understanding of data generated using them. Barccelona, Catalunya (Spain)

Jordi Llorens Marta Barenys

vii

Contents Preface to the Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

PART I

CYTOLOGICAL AND HISTOLOGICAL EVALUATION IN RODENTS

1 Amino-Cupric-Silver (A-Cu-Ag) Staining to Detect Neuronal Degeneration in the Mouse Brain: The de Olmos Technique . . . . . . . . . . . . . . . . . Rosario Moratalla, Adria´n Sanz-Magro, and Noelia Granado 2 Assessment of Auditory Hair Cell Loss by Cytocochleograms . . . . . . . . . . . . . . . . Aure´lie Thomas, Thomas Venet, and Benoıˆt Pouyatos 3 Evaluation of Cellular and Molecular Pathology in the Rodent Vestibular Sensory Epithelia by Immunofluorescent Staining and Confocal Microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Alberto F. Maroto, Erin A. Greguske, Alejandro Barrallo-Gimeno, and Jordi Llorens 4 Morphometric Analysis of Axons and Dendrites as a Tool for Assessing Neurotoxicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rhianna K. Morgan, Martin Schmuck, Ana Cristina Grodzki, Donald A. Bruun, Lauren E. Matelski, and Pamela J. Lein

PART II

v vii xiii

3 21

35

51

PHYSIOLOGICAL EVALUATION IN RODENTS

5 DPOAEs for the Assessment of Noise- or Toxicant-Induced Cochlear Damage in the Rat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 Thomas Venet, Aure´lie Thomas, Ludivine Wathier, and Benoıˆt Pouyatos 6 Use of Visual Evoked Potentials to Assess Deficits in Contrast Sensitivity in Rats Following Neurotoxicant Exposures . . . . . . . . . . . . . . . . . . . . . . 109 William K. Boyes 7 Electrophysiological Assessments in Peripheral Nerves and Spinal Cord in Rodent Models of Chemotherapy-Induced Painful Peripheral Neuropathy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 Susanna Park, Cynthia L. Renn, Justin G. Lees, Susan G. Dorsey, Guido Cavaletti, and Valentina A. Carozzi

PART III

BEHAVIORAL EVALUATION IN RODENTS

8 The Functional Observation Battery: Utility in Safety Assessment of New Molecular Entities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 David V. Gauvin

ix

x

Contents

9 Behavioral Assessment of Vestibular Dysfunction in Rats. . . . . . . . . . . . . . . . . . . . . 199 Alberto F. Maroto, Erin A. Greguske, Meritxell Deulofeu, Pere Boadas-Vaello, and Jordi Llorens 10 Assessment of Olfactory Toxicity in Rodents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217 David C. Dorman and Melanie L. Foster 11 Assessment of Neurotoxicant-Induced Changes in Behavior: Issues Related to Interpretation of Outcomes and Experimental Design . . . . . . . . . . . . . 239 Deborah A. Cory-Slechta, Katherine Harvey, and Marissa Sobolewski

PART IV 12

13

14

MOLECULAR METHODS

Assessment of Neurofilament Light Protein as a Serum Biomarker in Rodent Models of Toxic-Induced Peripheral Neuropathy . . . . . . . . . . . . . . . . . 267 Giulia Fumagalli, Guido Cavaletti, Henrik Zetterberg, and Cristina Meregalli Assessing Neurotoxicant-Induced Inflammation in the Central Nervous System: Cytokine mRNA with Immunostaining of Microglia Morphology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277 Christopher A. McPherson and G. Jean Harry Mitochondrial Stress Assay and Glycolytic Rate Assay in Microglia Using Agilent Seahorse Extracellular Flux Analyzers . . . . . . . . . . . . . . . . . . . . . . . . 305 Gabrielle Childers and G. Jean Harry

PART V ALTERNATIVE MODEL ORGANISM-BASED METHODS 15

16

17

Assessment of Larval Zebrafish Locomotor Activity for Developmental Neurotoxicity Screening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327 Bridgett N. Hill, Kayla D. Coldsnow, Deborah L. Hunter, Joan M. Hedge, David Korest, Kimberly A. Jarema, and Stephanie Padilla A Behavioral Test Battery to Assess Larval and Adult Zebrafish After Developmental Neurotoxic Exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353 Andrew B. Hawkey, Zade Holloway, and Edward D. Levin Evaluation of Neurotoxic Effects in Zebrafish Embryos by Automatic Measurement of Early Motor Behaviors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381 ¨ ver, Afolarin O. Ogungbemi, Eberhard Ku ¨ ster, Elisabet Teixido , Nils Klu and Stefan Scholz

Contents

18

xi

Application of Fluorescence Microscopy and Behavioral Assays to Demonstrating Neuronal Connectomes and Neurotransmitter Systems in C. elegans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399 Omamuyovwi M. Ijomone, Priscila Gubert, Comfort O. A. Okoh, ˜ o, Leandro de O. Amaral, Oritoke M. Aluko, Alexandre M. Vara and Michael Aschner

PART VI

IN VITRO METHODS

19

Effects of Neurotoxic or Pro-regenerative Agents on Motor and Sensory Neurite Outgrowth in Spinal Cord Organotypic Slices and DRG Explants in Culture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sara Bolı´var, Ilary Allodi, Mireia Herrando-Grabulosa, and Esther Udina 20 Integrative In Vitro/Ex Vivo Assessment of Dopaminergic Neurotoxicity in Rodents Using Striatal Synaptosomes and Membrane Preparations . . . . . . . . . Rau´l Lopez-Arnau and David Pubill 21 Quantification of Oligodendrocytes and Myelin in Human iPSC-Derived 3D Brain Cell Cultures (BrainSpheres) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . David Pamies, Megan Chesnut, He´le`ne Paschoud, Marie-Gabrielle Zurich, Thomas Hartung, and Helena T. Hogberg 22 Measurement of Electrical Activity of Differentiated Human iPSC-Derived Neurospheres Recorded by Microelectrode Arrays (MEA) . . . . . . . . . . . . . . . . . . . Kristina Bartmann, Julia Hartmann, Julia Kapr, and Ellen Fritsche Correction to: Application of Fluorescence Microscopy and Behavioral Assays to Demonstrating Neuronal Connectomes and Neurotransmitter Systems in C. elegans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

429

443

459

473

C1 489

Contributors ILARY ALLODI • Department of Neuroscience, University of Copenhagen, Copenhagen, Denmark ORITOKE M. ALUKO • The Neuro-Lab, School of Health and Health Technology, Federal University of Technology, Akure, Nigeria; Department of Physiology, School of Health and Health Technology, Federal University of Technology, Akure, Nigeria LEANDRO DE O. AMARA • Postgraduate Program in Pure and Applied Chemistry, Federal University of Western of Bahia, Bahia, Brazil MICHAEL ASCHNER • Departments of Molecular Pharmacology and Neurosciences, Albert Einstein College of Medicine, The Bronx, NY, USA ALEJANDRO BARRALLO-GIMENO • Departament de Cie`ncies Fisiolo`giques, Institut de Neurocie`nces, Universitat de Barcelona, Hospitalet de Llobregat, Catalunya, Spain; Institut d’Investigacio Biome`dica de Bellvitge, IDIBELL, Hospitalet de Llobregat, Catalunya, Spain KRISTINA BARTMANN • IUF-Leibniz Research Institute for Environmental Medicine, Duesseldorf, Germany PERE BOADAS-VAELLO • Research Group of Clinical Anatomy, Embryology and Neuroscience (NEOMA), Departament de Cie`ncies Me`diques, Facultat de Medicina, Universitat de Girona, Girona, Catalunya, Spain SARA BOLIVAR • Department of Cell Biology, Physiology and Immunology, Institute of Neurosciences, Centro de Investigacion Biome´dica en Red sobre Enfermedades Neurodegenerativas, Universitat Auto`noma de Barcelona, Bellaterra, Spain WILLIAM K. BOYES • Public Health and Integrated Toxicology Division, Center for Public Health and Environmental Assessment, Office of Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC, USA DONALD A. BRUUN • Department of Molecular Biosciences, School of Veterinary Medicine, University of California, Davis, Davis, CA, USA VALENTINA A. CAROZZI • Experimental Neurology Unit, Department of Medicine and Surgery, University of Milan-Bicocca, Milan, Italy; Young Against Pain Group, Milan, Italy GUIDO CAVALETTI • Experimental Neurology Unit, Department of Medicine and Surgery, University of Milan-Bicocca, Milan, Italy; Experimental Neurology Unit, School of Medicine and Surgery, and NeuroMI (Milan Center for Neuroscience), University of Milano-Bicocca, Monza(MB), Italy MEGAN CHESNUT • Center for Alternatives to Animal Testing (CAAT), Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA GABRIELLE CHILDERS • Neurotoxicology Group, National Toxicology Program Laboratory, National Institute of Environmental Health Sciences, Research Triangle Park, NC, USA; University of Alabama at Birmingham, Birmingham, AL, USA KAYLA D. COLDSNOW • Department of Biological Sciences, Rensselaer Polytechnic Institute, Troy, NY, USA DEBORAH A. CORY-SLECHTA • Department of Environmental Medicine, University of Rochester, Medical School, Rochester, NY, USA

xiii

xiv

Contributors

MERITXELL DEULOFEU • Research Group of Clinical Anatomy, Embryology and Neuroscience (NEOMA), Departament de Cie`ncies Me`diques, Facultat de Medicina, Universitat de Girona, Girona, Catalunya, Spain DAVID C. DORMAN • College of Veterinary Medicine, North Carolina State University, Raleigh, NC, USA SUSAN G. DORSEY • Department of Pain and Translational Symptom Science, School of Nursing, University of Maryland, Baltimore, MD, USA MELANIE L. FOSTER • Integrated Laboratory Systems LLC., Raleigh, NC, USA ELLEN FRITSCHE • IUF-Leibniz Research Institute for Environmental Medicine, Duesseldorf, Germany; Heinrich-Heine-University, Duesseldorf, Germany GIULIA FUMAGALLI • Experimental Neurology Unit, School of Medicine and Surgery, and NeuroMI (Milan Center for Neuroscience), University of Milano-Bicocca, Monza (MB), Italy DAVID V. GAUVIN • Neurobehavioral Sciences Department, Charles River Laboratories, Inc., Mattawan, MI, USA NOELIA GRANADO • Instituto Cajal (CSIC), Consejo Superior de Investigaciones Cientı´ficas, Madrid, Spain; CIBERNED, Instituto de Salud Carlos III, Madrid, Spain ERIN A. GREGUSKE • Departament de Cie`ncies Fisiolo`giques, Institut de Neurocie`nces, Universitat de Barcelona, Hospitalet de Llobregat, Catalunya, Spain; Institut d’Investigacio Biome`dica de Bellvitge, IDIBELL, Hospitalet de Llobregat, Catalunya, Spain ANA CRISTINA GRODZKI • Department of Molecular Biosciences, School of Veterinary Medicine, University of California, Davis, Davis, CA, USA PRISCILA GUBERT • Department of Biochemistry, Laboratorio de Imunopatologia Keizo Asami, LIKA, Federal University of Pernambuco, Recife, Brazil; Postgraduate Program in Pure and Applied Chemistry, Federal University of Western of Bahia, Bahia, Brazil G. JEAN HARRY • Neurotoxicology Group, National Toxicology Program Laboratory, National Institute of Environmental Health Sciences, Research Triangle Park, NC, USA JULIA HARTMANN • IUF-Leibniz Research Institute for Environmental Medicine, Duesseldorf, Germany THOMAS HARTUNG • Center for Alternatives to Animal Testing (CAAT), Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA; Center for Alternatives to Animal Testing-Europe, University of Konstanz, Konstanz, Germany; AxoSim, Inc., New Orleans, LA, USA KATHERINE HARVEY • Department of Environmental Medicine, University of Rochester, Medical School, Rochester, NY, USA ANDREW B. HAWKEY • Department of Psychiatry and Behavioral Sciences, Duke University Medical Center, Durham, NC, USA JOAN M. HEDGE • Office of Research and Development, Center for Computational Toxicology and Exposure, U.S. Environmental Protection Agency, Research Triangle Park, NC, USA MIREIA HERRANDO-GRABULOSA • Department of Cell Biology, Physiology and Immunology, Institute of Neurosciences, Centro de Investigacion Biome´dica en Red sobre Enfermedades Neurodegenerativas, Universitat Auto`noma de Barcelona, Bellaterra, Spain BRIDGETT N. HILL • Oak Ridge Institute for Science and Education, Oak Ridge, TN, USA HELENA T. HOGBERG • Center for Alternatives to Animal Testing (CAAT), Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA ZADE HOLLOWAY • Department of Psychiatry and Behavioral Sciences, Duke University Medical Center, Durham, NC, USA

Contributors

xv

DEBORAH L. HUNTER • Office of Research and Development, Center for Computational Toxicology and Exposure, U.S. Environmental Protection Agency, Research Triangle Park, NC, USA OMAMUYOVWI M. IJOMONE • The Neuro-Lab, School of Health and Health Technology, Federal University of Technology, Akure, Nigeria; Department of Human Anatomy, School of Health and Health Technology, Federal University of Technology, Akure, Nigeria KIMBERLY A. JAREMA • Office of Research and Development, Center for Public Health and Environmental Assessment, U.S. Environmental Protection Agency, Research Triangle Park, NC, USA JULIA KAPR • IUF-Leibniz Research Institute for Environmental Medicine, Duesseldorf, Germany NILS KLU¨VER • Department of Bioanalytical Ecotoxicology, Helmholtz Centre for Environmental Research—UFZ, Leipzig, Germany DAVID KOREST • Oak Ridge Institute for Science and Education, Oak Ridge, TN, USA EBERHARD KU¨STER • Department of Bioanalytical Ecotoxicology, Helmholtz Centre for Environmental Research—UFZ, Leipzig, Germany JUSTIN G. LEES • School of Medical Sciences, The University of New South Wales, Sydney, Australia PAMELA J. LEIN • Department of Molecular Biosciences, School of Veterinary Medicine, University of California, Davis, Davis, CA, USA EDWARD D. LEVIN • Department of Psychiatry and Behavioral Sciences, Duke University Medical Center, Durham, NC, USA JORDI LLORENS • Departament de Cie`ncies Fisiolo`giques, Institut de Neurocie`nces, Universitat de Barcelona, Hospitalet de Llobregat, Catalunya, Spain; Institut d’Investigacio Biome`dica de Bellvitge, IDIBELL, Hospitalet de Llobregat, Catalunya, Spain RAU´L LO´PEZ-ARNAU • Faculty of Pharmacy and Food Sciences, Department of Pharmacology, Toxicology and Therapeutic Chemistry, Pharmacology Section and Institute of Biomedicine (IBUB), Universitat de Barcelona, Barcelona, Spain ALBERTO F. MAROTO • Departament de Cie`ncies Fisiolo`giques, Institut de Neurocie`nces, Universitat de Barcelona, Hospitalet de Llobregat, Catalunya, Spain; Institut d’Investigacio Biome`dica de Bellvitge, IDIBELL, Hospitalet de Llobregat, Catalunya, Spain LAUREN E. MATELSKI • Department of Molecular Biosciences, School of Veterinary Medicine, University of California, Davis, Davis, CA, USA CHRISTOPHER A. MCPHERSON • Neurotoxicology Group, National Toxicology Program Laboratory, National Institute of Environmental Health Sciences, Research Triangle Park, NC, USA CRISTINA MEREGALLI • Experimental Neurology Unit, School of Medicine and Surgery, and NeuroMI (Milan Center for Neuroscience), University of Milano-Bicocca, Monza (MB), Italy ROSARIO MORATALLA • Instituto Cajal (CSIC), Consejo Superior de Investigaciones Cientı´ficas, Madrid, Spain; CIBERNED, Instituto de Salud Carlos III, Madrid, Spain RHIANNA K. MORGAN • Department of Molecular Biosciences, School of Veterinary Medicine, University of California, Davis, Davis, CA, USA AFOLARIN O. OGUNGBEMI • Department of Bioanalytical Ecotoxicology, Helmholtz Centre for Environmental Research—UFZ, Leipzig, Germany COMFORT O. A. OKOH • The Neuro-Lab, School of Health and Health Technology, Federal University of Technology, Akure, Nigeria

xvi

Contributors

STEPHANIE PADILLA • Office of Research and Development, Center for Computational Toxicology and Exposure, U.S. Environmental Protection Agency, Research Triangle Park, NC, USA DAVID PAMIES • Center for Alternatives to Animal Testing (CAAT), Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA; Department of Biomedical Science, University of Lausanne, Lausanne, Switzerland; Swiss Centre for Applied Human Toxicology (SCAHT), Lausanne, Switzerland SUSANNA PARK • Brain and Mind Centre, Faculty of Medicine and Health, University of Sydney, Sydney, Australia HE´LE`NE PASCHOUD • Department of Biomedical Science, University of Lausanne, Lausanne, Switzerland BENOIˆT POUYATOS • National Research and Safety Institute for the Prevention of Occupational Accidents and Diseases (INRS), Vandoeuvre-le`s-Nancy, France DAVID PUBILL • Faculty of Pharmacy and Food Sciences, Department of Pharmacology, Toxicology and Therapeutic Chemistry, Pharmacology Section and Institute of Biomedicine (IBUB), Universitat de Barcelona, Barcelona, Spain CYNTHIA L. RENN • Department of Pain and Translational Symptom Science, School of Nursing, University of Maryland, Baltimore, MD, USA ADRIA´N SANZ-MAGRO • Instituto Cajal (CSIC), Consejo Superior de Investigaciones Cientı´ficas, Madrid, Spain; CIBERNED, Instituto de Salud Carlos III, Madrid, Spain MARTIN SCHMUCK • Department of Molecular Biosciences, School of Veterinary Medicine, University of California, Davis, Davis, CA, USA STEFAN SCHOLZ • Department of Bioanalytical Ecotoxicology, Helmholtz Centre for Environmental Research—UFZ, Leipzig, Germany MARISSA SOBOLEWSKI • Department of Environmental Medicine, University of Rochester, Medical School, Rochester, NY, USA ELISABET TEIXIDO´ • Department of Bioanalytical Ecotoxicology, Helmholtz Centre for Environmental Research—UFZ, Leipzig, Germany; GRET-Toxicology Unit, Faculty of Pharmacy and Food Sciences, Department of Pharmacology, Toxicology and Therapeutic Chemistry, University of Barcelona, Barcelona, Spain AURE´LIE THOMAS • National Research and Safety Institute for the Prevention of Occupational Accidents and Diseases (INRS), Vandoeuvre-le`s-Nancy, France ESTHER UDINA • Department of Cell Biology, Physiology and Immunology, Institute of Neurosciences, Centro de Investigacion Biome´dica en Red sobre Enfermedades Neurodegenerativas, Universitat Auto`noma de Barcelona, Bellaterra, Spain ALEXANDRE M. VARA˜O • Postgraduate Program in Pure and Applied Chemistry, Federal University of Western of Bahia, Bahia, Brazil THOMAS VENET • National Research and Safety Institute for the Prevention of Occupational Accidents and Diseases (INRS), Vandoeuvre-le`s-Nancy, France LUDIVINE WATHIER • National Research and Safety Institute for the Prevention of Occupational Accidents and Diseases (INRS), Vandoeuvre-le`s-Nancy, France HENRIK ZETTERBERG • Department of Psychiatry and Neurochemistry, The Sahlgrenska Academy at the University of Gothenburg, Mo¨lndal, Sweden; Clinical Neurochemistry Laboratory, Sahlgrenska University Hospital, Mo¨lndal, Sweden; Department of Neurodegenerative Disease, UCL Institute of Neurology, Queen Square, London, UK; UK Dementia Research Institute at UCL, London, UK MARIE-GABRIELLE ZURICH • Department of Biomedical Science, University of Lausanne, Lausanne, Switzerland; Swiss Centre for Applied Human Toxicology (SCAHT), Lausanne, Switzerland

Part I Cytological and Histological Evaluation in Rodents

Chapter 1 Amino-Cupric-Silver (A-Cu-Ag) Staining to Detect Neuronal Degeneration in the Mouse Brain: The de Olmos Technique Rosario Moratalla, Adria´n Sanz-Magro, and Noelia Granado Abstract Silver staining procedures have classically been used to study the structure of the nervous system. However, reduced silver staining methods, using substances that reduce silver ions against the natural reducing properties of the tissue, can also successfully reveal degenerative changes in the nervous system. It is not known how silver binds these degenerating elements. During degeneration, silver ions may form complexes with exposed amino acid chains in denatured proteins that are then seen as black-stained elements over an unstained background (of non-degenerating elements). The reduced Amino-Cupric-Silver method developed by de Olmos, where cupric ions are added as an external reducer, provides the greatest contrast between the degenerating and non-degenerating neuronal elements when compared with other reduced silver staining protocols. This method is unique to study degenerative morphological changes and, when combined with other staining procedures, to identify which specific neuronal population is degenerating. After a trauma, neurons undergo several physical changes that can be visualized with different markers, such as FluoroJade, caspases, or Hematoxylin and Eosin stains. Some neurons can overcome this damage and are restored, while others undergo irreversible changes and die. The Amino-Cupric-Silver method can reveal early irreversible neuronal damage before cell death. After these irreversible changes the neurons cannot regenerate, so detecting these degenerative changes will improve the understanding of pathological changes following certain injuries of the nervous system. Key words Reduced silver stain procedures, A-Cu-Ag method, Neurotoxicity, Neurodegeneration, Cell death

1

Introduction Cognitive function depends on the correct activity and integrity of neurons and their pathways. The loss of a neuron is the last step of a degenerative process that follows certain types of damage [1]. As a consequence of the plasticity of neural systems and their compensatory capacity, neurological symptoms frequently appear only when neuron death is advance, and the system has lost a significant

Adria´n Sanz-Magro and Noelia Granado contributed equally to this work. Jordi Llorens and Marta Barenys (eds.), Experimental Neurotoxicology Methods, Neuromethods, vol. 172, https://doi.org/10.1007/978-1-0716-1637-6_1, © Springer Science+Business Media, LLC, part of Springer Nature 2021

3

4

Rosario Moratalla et al.

number of cells [2]. Also, considering that neurons cannot regenerate, it is essential to identify early irreversible neuronal damage [1]. Detecting degenerative changes before neuron death is a challenge in the field of neurotoxicology because it requires predicting where the neuronal loss will take place as a consequence of a certain process [1]. Based on the affinity of nervous system cells for silver (argyrophilia), silver staining procedures were largely used in neuroanatomy during the last century to study the structure of the nervous system. These methods were successfully employed to stain a large variety of neuronal elements and trace axonal pathways [1, 3, 4]. One of the first silver stain method was designed by Bielchowsky in 1904 [4]. This method stained all type of neural cells so, it was difficult to differentiate between normal and pathological neuronal elements [1]. Later on, Nauta and Gygax, improved this method introducing a pretreatment with phosphomolybdic acid-potassium permanganate solution, achieving a better contrast between normal and damaged elements [1, 5]. Although this modification can reveal neuronal degeneration [1, 3], the development of reduced silver methods further improved it, reaching its maximal strength and providing the optimal tool to approach neurodegeneration studies [1, 3]. With the reduced silver staining methods (in which silver ions are reduced to metallic silver against the natural silver-reducing properties of the nervous tissue), the degenerating neuronal elements are more intensely stained while non-degenerating elements (background) remain unstained [1, 3, 6, 7]. Despite the usefulness of these methods to study nervous system, the rise of new stains (e.g., horseradish peroxidase, fluorescent tracer methods) has caused them to be less commonly used [1, 3]. The main mechanism explaining how staining results from the interaction between the silver and the degenerating elements is still unclear [1, 8]. The most widely accepted hypothesis is that silver ions react with exposed single amino acids or short amino acid sequences [9, 10] and form molecular associations of a central metal cation (silver) surrounded by a certain number of amino acid ions [11, 12]. In a healthy cell, proteins have a globular conformation and amino acid groups are not exposed to react with silver ions. In a damaged cell, proteolytic mechanisms can alter protein conformations and expose these residues, which can then easily react with silver ions [1, 3, 8, 13]. The first known reduced silver staining is the “Golgi Stain,” developed by Camillo Golgi in 1873, using potassium dichromate as a reducer. Santiago Ramo´n y Cajal used the Golgi Stain with small modifications to study the structure of the nervous system and could validate his “neuron doctrine” [1, 14]. Later on, de Olmos developed the reduced Amino-Cupric-Silver (A-Cu-Ag) method [6, 7, 15], where cupric ions decrease the argyrophilia of

Amino Cupric Silver (A-Cu-Ag) Staining Neurotoxicity

5

non-degenerating elements, but not in degenerating ones, providing the strongest contrast until present between degenerating and non-degenerating elements [1, 3]. The A-Cu-Ag method is based on the following steps: (1) pretreatment with a cupric solution that provides specificity to the staining of degenerating elements [3, 6, 7]; (2) silver impregnation of the samples with a silver nitrate (AgNO3) solution; (3) reduction of silver ions to metallic silver around exposed amino acid groups in the tissue, forming microscopically visible nuclei; (4) bleaching to clear the background and increase the contrast between normal and degenerating elements; and (5) stabilization of the staining [1, 7]. Because silver can stain neuronal elements at different degenerative stages [7], the A-Cu-Ag method can be used not only to identify degeneration, but also to study the progression of degeneration, through the accumulation of silver deposits over time and the loss of the corresponding degenerating neurons or fibers associated with specific pathological changes, as shown previously [1, 7, 10, 16], see also Fig. 1. During a degenerative process, dendrites and axonal processes start to fragment in rows with slight silver staining, followed by the appearance of punctuate structures, and then their total disappearance [7]. Generally, in degeneration, cell somas undergo changes in size and shape, at first without the presence of silver deposits. As the degenerating soma get smaller the silver starts to accumulate. The neuron can then became dysfunctional and start losing the capacity to synthesize their phenotypic and domestic markers. Finally, small silver-stained debris remains until the soma completely disappears [1, 7, 13], see also Fig. 2. With methamphetamine, all these pathological changes can be observed as soon as 24 h after the administration (Fig. 2) although these changes do not necessarily represent sequential steps of the degenerating process, due to the severity of the insult. Nevertheless, we do observe all these degenerating changes [13]. Furthermore, as illustrated in Figs. 2 and 3, and mentioned above, because of the unstained background, the A-Cu-Ag method can be successfully combined with other stains [13, 17]. This approach allows the identification of specific types of degenerating neurons [13, 18–20]. Previous studies by our group, combining the A-Cu-Ag method with immunohistochemistry against the enzyme thyroxin hydroxylase (TH), a marker of dopaminergic neurons, or with green fluorescent protein or red-tomato protein, demonstrated that methamphetamine induces degeneration in dopaminergic cells of the substantia nigra pars compacta (SNpc) and in its projections to the striatum [13, 18]. Indeed, we showed that the peak of silver deposits after methamphetamine overlaps with that of striatal TH-immunoreactivity (TH-ir) fiber loss (Figs. 1 and 3) and furthermore, that silver deposits also occurred in the soma of dopaminergic neurons in the SNpc (Fig. 2).

6

Rosario Moratalla et al.

Fig. 1 Methamphetamine induces a progressive TH-ir fiber degeneration in the mice striatum. Photomicrographs of striatal coronal sections of mice treated with saline or methamphetamine (METH). Sections are stained against TH and adjacent section with the A-Cu-Ag method. Mice were sacrificed 3, 12 or 24 h after METH. Note the progression of dopaminergic terminal loss demonstrated by the decrease in TH-ir and its correlation with the progressive increase of silver deposits shown in adjacent sections. Histograms illustrate the quantification analysis, showing the proportional stained area with TH-ir or with the A-Cu-Ag method in the striatum. Data represent the mean
SEM, n ¼ 4–6 per group. Scale bar indicates 500 μm

An important factor to consider in the A-Cu-Ag method is the time between injury and when the stain is performed [1, 7]. Degenerative changes appear and remain at different times depending on the neuronal compartment where they are taking place or on the type of insult or toxin used [1, 7]. In our experiments with

Amino Cupric Silver (A-Cu-Ag) Staining Neurotoxicity

7

Fig. 2 Different stages of degeneration in dopaminergic neurons in the SNpc after methamphetamine. Photomicrographs of double TH/A-Cu-Ag-stained neurons of the SNpc of mice sacrificed 24 h after methamphetamine. Note that some degenerating neurons change their size and shape (a) and have silver deposits in their soma (b and c) but still express the dopaminergic marker, TH in this case. Also, neurons with only silver deposit indicating the irreversible degeneration process (d) and neurons with silver-stained debris that have already started to disappear (e–f). Scale bar indicates 20 μm

methamphetamine, silver deposits in dopaminergic terminals in the striatum appear few hours after treatment (around 3 h) and are still evident 3–7 days after the insult (Fig. 1) [13]. However, in the dopaminergic cell bodies of the SNpc, silver deposits appear later than in terminals (around 24 h) and can remain around 3–5 days [13]. Considering this, the best time to detect degenerative changes in the dopaminergic nigrostriatal pathway caused by methamphetamine administration is around 24 h after the insult because at this time the damage appears in both, somas and synaptic terminals, and it is also when it reaches the peak [13]. However, it must be taken into consideration that the appearance and duration of the degenerative changes could differ depending on the type of insult and the type of the neuronal target [1, 7]. Also, the A-Cu-Ag method is an ideal tool to anatomically localize neurodegeneration, allowing to determine its anatomical pattern [1, 7, 13]. In our experiments, we observed that dopaminergic degeneration in the striatum caused by methamphetamine is not homogeneous, and the same happens with the A-Cu-Ag

8

Rosario Moratalla et al.

Fig. 3 Double A-Cu-Ag and TH staining. Photomicrographs of the coronal sections of single and double TH/ACu-Ag-stained sections of the striatum of mice sacrificed 24 h after methamphetamine. Because of the clear background of the A-Cu-Ag method, degenerating elements and residual TH-ir fibers can be observed in the same sections in the striatum. Scale bar indicates 500 μm

staining. Indeed, there is an overlap pattern of TH fibers loss and silver deposits (Figs. 1 and 3). Furthermore, we could even observe small striatal areas with greater TH-ir fiber loss that overlaps with greater silver deposits (Fig. 4). Curiously, these areas corresponded in shape and in numbers with striatal striosomes as we could demonstrate using μ-opioid receptor (MOR-1) as a marker for the striosomal compartment [21]. Despite the advantages of the A-Cu-Ag method [3, 6, 7, 15, 16], the development of new faster and more simple staining methods, that can indirectly show damage in the nervous system, has decreased its use. However, this A-Cu-Ag method can be complemented with other methods [13], such as the FluoroJade stain, that detects alterations in the cytoplasm or in the nucleus of degenerating neurons or hematoxylin and eosin (H&E) stain that detect eosinophilic necrotic neurons [1, 13, 22]. Nissl stain, that detects apoptotic cell bodies, can also be used in combination with the silver method as have been successfully shown in SNpc after methamphetamine [13, 23]. As inflammatory responses often accompany degeneration, specific immunohistochemistry assays such against astroglia and microglia markers can also detect degeneration when combined in adjacent section with the silver method as have shown before [18, 21]. DNA damage proteins or caspase activation can also provide information about the response to an injury [16, 23–27]. Nevertheless, all these methods fail to detect structural and pathological changes in small degenerating structures such as synaptic terminals, dendrites, and axons [1], but the A-Cu-Ag, due to the described characteristics, allows to successfully

Amino Cupric Silver (A-Cu-Ag) Staining Neurotoxicity

9

Fig. 4 Correlation between terminal degeneration and silver deposits: Striosomes are more vulnerable to methamphetamine and show higher silver deposits. Photomicrographs of adjacent striatal coronal sections stained against TH or with the A-Cu-Ag method. Mice were sacrificed 24 h after methamphetamine administration. Note that striatal striosomes undergo stronger degeneration after methamphetamine and show denser silver deposits. Arrows indicate striatal striosomes and boxes the magnified area. Scale bar indicates 500 μm at lower magnification and 250 μm at high magnification

studying these changes. Many studies have used the A-Cu-Ag method to specifically identify neuronal damage induced by neurotoxins [7, 17, 28, 29], drugs of abuse [13, 18–20, 30], physical damage [7, 31], or altered physiological processes [6, 7]. In summary, because of features such as (1) specificity, (2) power of contrast, and (3) the possibility to combine with other staining techniques, the A-Cu-Ag method may provide useful information on the degenerative changes after an injury to the nervous system. The main goal of this chapter is to provide a comprehensive and simple protocol for the A-Cu-Ag method so it can be replicated in other laboratories. Besides the A-Cu-Ag method, we include a TH immunohistochemistry protocol that could be carry out in adjacent or even in the same sections that the A-Cu-Ag method, in order to gain anatomical depth and resolution and to identify the phenotype of the degenerating elements. We also provide protocols to quantify

10

Rosario Moratalla et al.

degeneration, studying fiber loss or silver deposits by measuring the proportional stained area or optical density and a stereological protocol to quantify the number of degenerating neurons (cell soma with silver deposits). The protocol described below successfully detects dopaminergic degeneration after methamphetamine administration in the striatum and SNpc of mice of either sex, or different ages, even embryos or neonates and can be done in coronal, sagittal, or horizontal brain sections. Control brain sections from animals treated with saline should be processed at the same time with the experimental sections to serve as reference for the bleaching and other steps and for quantification studies. The conditions described in this chapter are adapted to detect dopaminergic damage in the striatum and SNpc caused by methamphetamine administration. The experimenter is advised to test different conditions in order to obtain the optimal results with the specific insult under study.

2

Materials

2.1 Tissue Fixation and Brain Preparation

Subjects and drugs: This protocol uses 3–4 months old C57BL/6J mice. To induce dopaminergic degeneration, we use methamphetamine hydrochloride (SIGMA ref M-8750) dissolved in 0.9% sodium chloride (NaCl or saline; MERCK ref 1.06404.1000) given in three doses of 5 mg/kg at 3 h intervals. This dose regimen causes a clear degeneration in the brain, as previously demonstrated [23–25]. Animals are sacrificed at different time points after methamphetamine treatment (3, 12, and 24 h after methamphetamine administration), in order to assess different time points of dopamine degeneration in the striatum and SNpc. Before start with the protocol, carefully read Notes 1–4. Perfusion: The following solutions are needed to perform the perfusion process: l

Anesthetic: Sodium pentobarbital ( VETOQUINOL ; 50 mg/ kg).

l

Rinse solution: 0.8% NaCl (MERCK ref 1.06404.1000), 0.4% glucose (SIGMA ref G7021), 0.8% sucrose (MERCK, ref 107687), and 0.5% sodium nitrite (NaNO2; SIGMA ref S2252), as a vasodilator, in bi-distilled water (H2Obd).

l

Fixation solution (paraformaldehyde; PFA): 4% PFA (SIGMA ref 441244) with 50 mg sodium sulfite (Na2SO3; SIGMA ref. S0505) per L of H2Obd; pH 7.6–7.8 (see Note 5).

l

0.2 M borate buffer pH 8.5: 50 mg Na2SO3 (SIGMA ref S2252) and 16.8 mg sodium tetraborate (Borax; SIGMA ref 71997) in 1 L H2Obd (light sensitive).

Amino Cupric Silver (A-Cu-Ag) Staining Neurotoxicity

11

Tissue processing: For rodent brains, cut 50 μm thick sections (see Note 6) with a Vibratome (Leica Microsystems GmbH, Wetzlar, Germany) on free-floating way using 0.2 M borate buffer, pH 8.5. 2.2 The A-Cu-Ag Method

Pre-impregnation solution Prepare 100 mL of the pre-impregnation solution by combining the following reagents in this order: 1. 100 mg silver nitrate (AgNO3; SIGMA ref 85228; light sensitive; never weigh with metallic instruments). 2. 100 mL H2Obd. 3. 53 mg dl-α-amino-n-butyric acid (SIGMA ref 162663). 4. 46 mg dl-alanine (SIGMA ref A7502). 5. 2 mL 0.5% copper nitrate (Cu (NO3)2; SIGMA ref 229636; light sensitive). 6. 0.2 mL 0.5% cadmium nitrate (Cd(NO3)2; SIGMA ref 20911). 7. 1.5 mL 0.5% lanthanum nitrate (La(NO3)3; MERCK ref 2506). 8. 0.5 mL 0.5% Neutral Red (FLUKA ref 72210). 9. 1.0 mL pyridine (MERCK ref 1.097828.0500; toxic, always work in a fume hood and with the proper respiratory protection. 10. 1.0 mL triethanolamine (SIGMA ref T-58300). 11. 2.0 mL isopropanol (MERCK ref 1.09634.1011; work in a fume hood. Impregnation solution Combine the following compounds: 1. 5 mL H2Obd. 2. 412 mg AgNO3 (SIGMA ref 85228; completely dissolved before adding the 100% EtOH). 3. 4 mL 100% EtOH (MERCK ref 1.00983.2511). 4. 50 μL acetone (MERCK ref 1.00014.1011). 5. 3 mL 0.4% lithium hydroxide (LiOH; SIGMA ref I-4533; store at 4  C). 6. 0.65 mL ammonium hydroxide (NH4OH; FISHER ref 10552174; toxic, always work in a fume hood, use within 2–3 months after preparation). Reduction solution: This solution can be prepared several days before use. Store at room temperature and protect from light. To prepare 1 L of reduction solution, combine:

12

Rosario Moratalla et al.

1. 800 mL distilled water (H2Od). 2. 90 mL 100% EtOH (MERCK ref 1.00983.2511). 3. 11 mL 10% formalin (SIGMA ref 47608; store at 4  C). 4. 6.5 mL 1% citric acid monohydrate (FISHER ref 16345410; freshly prepared). Bleaching solution: Two bleaching solutions are required. l

Solution 1: prepared in lactic acid (MERK ref 1.00366.2500) – 6% potassium ferricyanide (K3[Fe(CN)6]; MERCK ref 1.04973.0250). – 4% potassium 131493.1210).

l

chlorate

(KClO3;

PANREAC

ref

Solution 2: prepared in H2Od – 0.06% potassium permanganate (KMnO4; provided by Dr. R. Martı´nez, Cajal Institute, CSIC). – 5% sulfuric acid (FISHER ref S/9240/PB15). Stabilization solution: Two stabilization solutions are required:

2.3 TH Immunohistochemistry

l

Stabilization solution: 2% Sodium thiosulfate (Na2S2O3; SIGMA ref 217247) in H2Od.

l

Rapid Fixer solution: KODAK concentrated Rapid Fixer solution A + B (SIGMA ref P7542-1GA), diluted 1:6 in H2Od.

For immunohistochemistry, prepare the following buffers and solutions: l

0.4 M phosphate buffer (PB) pH 7.4: 27.6 g sodium phosphate monobasic monohydrate (NaH2PO4∙H2O; SIGMA ref S-9638), and 106 g sodium phosphate dibasic (Na2HPO4∙2H2O; MERCK ref 1.06580.1000) in 2 L H2Od.

l

0.1 M PB: 500 mL PB4 in 1.5 L H2Od.

l

Phosphate-buffered saline (PBS) pH 7.4: 500 mL PB4 and 18 g NaCl in 1.5 L H2Od.

l

PBS-TX: 0.2% Triton X-100 (SIGMA ref T8787) in PBS buffer.

l

3% peroxide (H2O2; PANREAC ref 141076.1211) in PBS-TX.

l

10% Normal Goat Serum (NGS; MERCK ref S26-100ML) in PBS-TX.

l

TH primary antibody: 1:1000 rabbit anti-TH (MERCK-MILLI PORE ref AB152) in PBS-TX, 1% NGS.

l

Secondary antibody: 1:500 goat anti-rabbit biotinylated antibody (Vector Labs ref BA1000) in PBS-TX, 1% NGS.

Amino Cupric Silver (A-Cu-Ag) Staining Neurotoxicity l

l

2.4 Mounting and Visualization

2.5 Image Analysis Quantification and Stereology

13

1:5.000 Streptavidin (INVITROGEN ref 434323) in PBS-TX, 1% NGS. 3,30 -Diaminobenzidine (DAB; SIGMA ref D5637; light sensitive; toxic; 50 mg/100 mL of 0.1 M PB).

l

Gelatin solution for slides: add 1.5 g gelatin to 50 mL hot H2Od, wait until dissolved, then add 50–80 mL of ethanol.

l

Xylene (VWR ref. 141769.2711).

l

DPX Mounting Medium (MERCK ref 1.00579.0500).

l

Microscope: to visualize the degenerating elements (both striatum and SNpc), we use an optic microscope equipped with 4 lens and a Leica DFC 290 HD video camera to acquire the micro-pictures.

l

Image analysis system: to study TH-ir and the silver impregnation, we use the Image J program (National Institutes of Health, Bethesda, Maryland, USA) that converts color intensities into a gray color scale. This allows us to quantify the stained area as the proportion of staining pixels in relation to the total pixels in the selected area [32].

To count the number of degenerating and dopaminergic cell bodies, we use a stereological software (see Subheading 3.5) coupled to a Nikon Eclipse 80i microscope. The microscope is connected to (1) an interactive computer system comprising a high-precision motorized microscope stage, (2) a 0.5 μm resolution microadaptor (Heidenhain VZR401), (3) a solid-state Microbrightfield CX9000 videocamera, and (4) a high-resolution video monitor using the optical fractionator Stereo Investigator program (Microbrightfield Bioscience, Colchester, VT) [33].

3

Methods The following A-Cu-Ag method protocol is taken from the one described by de Olmos et al. 1994 [7].

3.1 Tissue Fixation and Preparation

Good tissue fixation is a critical step in the A-Cu-Ag method (perfusion to remove all blood elements from the tissue and avoid a less clear background) [6, 16]. For the fixation solution (see Subheading 2.1), heat H2Obd in a glass cup to 60  C. Add Na2SO3 and a small amount of Borax (4 g for 1 L of solution). When the Borax is completely dissolved, add the PFA and stir at 50  C until completely dissolved (10 mg boric acid per 1 L of solution). Allow the solution to cool for 45–60 min and adjust to a pH of 7.45 (see Note 5).

14

Rosario Moratalla et al.

Anesthetize the animals (see Subheading 2.1) with sodium pentobarbital and perfuse transcardially with the fixation solution (PFA, see Subheading 2.1). Before and after the fixation, perfuse the animal with 10 mL (around one-third of the animal’s weight) of the rinse solution (see Subheading 2.1). As a vasodilator, use 0.5% sodium nitrite. Leave brains overnight in the skull at 4  C to avoid any extra damage during dissection that will be detected by the silver stain, then remove the brain and store in 30% sucrose for 2 days. Place the brains in a solution of 3% agarose in 0.2 M borate buffer, pH 8.5, and cut into 50 μm thick coronal sections in a vibratome. Store in 4% PFA (see Note 7). 3.2 The A-Cu-Ag Method

Pre-impregnation: During pre-impregnation, the addition of cupric ions will selectively stain degenerating neuronal elements (see Subheading 2.2 and Notes 8 and 9) [3, 6, 7, 15, 17]. Before use, heat the pre-impregnation solution in the microwave to 48  C. If the solution has precipitated, filter with glass fiber filters. Take the selected slices from the 4% PFA and wash with H2Obd in a porcelain filter. Put the slices into the pre-impregnation solution and heat in the microwave to 47–48  C. Allow to cool at room temperature for 2–3 h. Tissue must turn brown. Impregnation: Take slices from the pre-impregnation solution and place in a porcelain filter. Wash for 1 min with H2Obd and twice with acetone (45 s–1 min for each wash) to remove excess pyridine and pre-impregnation solution. Transfer the slices into the impregnation solution (see Subheading 2.2) for 45–50 min (see Note 10). Reduction: Heat the reduction solution (see Subheading 2.2) to 30  C. Transfer the slices from the impregnation solution to the reduction solution and incubate for 20 min (see Note 11). Tissue must get darker. Wash the slices in H2Od and place into 0.25% acetic acid for 1 min to stop the reduction reaction (do not move vigorously as the tissue can easily break). Perform two washes of 1 min with H2Obd and keep the slices at 4  C for at least 1 h before the bleaching treatment. Bleaching is performed in two steps (see Note 12): 1. Put sections in the first bleaching solution (see Subheading 2.2) in a porcelain desiccator at room temperature until they become relatively transparent (approximately 60 s) to remove background silver deposits but not degenerating elements, then wash sections with H2Od. 2. Transfer slices to the second bleaching solution (see Subheading 2.2) in a porcelain desiccator. Leave until the slices acquire a yellow color, around 15–20 s (if longer, the silver fixed in the

Amino Cupric Silver (A-Cu-Ag) Staining Neurotoxicity

15

degenerating elements could disappear). Put slices into H2Od and perform two washes of 5 min each before the stabilization step (see Note 13). Stabilization: Transfer sections to the stabilization solution (see Subheading 2.2) for approximately 1 min (in slow agitation). Sections should lose their yellow color and become transparent. Wash with H2Od for 5 min and put slices in the Rapid Fixer solution (see Subheading 2.2) for 1 min. Finally, put slices in H2Od. 3.3 TH Immunohistochemistry

Perform the immunohistochemistry on free-floating sections previously stained with the A-Cu-Ag method [13, 18]. Also, immunohistochemistry can be done alone in adjacent sections to the ones stained with the A-Cu-Ag method. After stabilization, wash the slices in 0.2% PBS-TX (see Subheading 2.3). Quench endogenous peroxidase incubating slices in 3% H2O2 in 0.2% PBS-TX for 10 min. Wash again in PBS-TX and block nonspecific binding sites for 60–90 min with 10% normal goat serum in 0.2% PBS-TX. Incubate the primary antibody antiTH (see Subheading 2.3) at room temperature overnight. Wash slices with 0.2% PBS-TX and incubate with the goat anti-rabbit biotinylated antibody (see Subheading 2.3) for 1–2 h at room temperature. Wash with 0.2% PBS-TX, incubate sections in streptavidin (see Subheading 2.3) for 1 h, and then stain with DAB.

3.4 Mounting and Visualization

After the DAB reaction, carefully mount the slices in gelatinized slides (see Subheading 2.4) [34]. When the tissue is completely dry, dehydrate the section through the following alcohol battery (10 min per solution): H2Od, 70% ethanol, 96% ethanol, 100% ethanol, 100% ethanol, and xylene. Cover the slices with DPX as the mounting medium and dry them for around 24 h before visualization under a microscope.

3.5 Image Analysis Quantification and Stereology

Quantitative assessment of degenerating dopaminergic terminals or silver deposits is done by evaluating the proportional stained area or by optical density. First, take pictures of the region of interest (striatum, in this case) with a 4 lens in an optical microscope (see Note 14). Pictures should be taken using the same light conditions and filters in the microscope for all the samples. Analyze pictures using an ImageJ (see Subheading 2.5) as previous described [13, 18–20] (Fig. 1). There are many free programs available; one of the most robust and widely used programs is the Image J. Stereological analysis estimate the number of degenerating cell somas (in our case, dopaminergic neurons), by counting the number of A-Cu-Ag-stained neurons in the SNpc using the optical fractionator Stereo Investigator program (Microbrightfield Bioscience, Colchester, VT), coupled to a Nikon Eclipse 80i microscope (see Subheading 2.4) (see Note 15). To assess methamphetamine

16

Rosario Moratalla et al.

toxicity in dopaminergic neurons in mice, around 4–6 animals per group are needed to obtain significant values. First, draw the outline of the SNpc (see Note 16) with a 2 lens in one out of four serial adjacent sections through the SNpc (usually around 9–12 sections per animal). Second, count the cells at higher magnification (60 or 100). At this high magnification, cell somas can be identified as A-Cu-Ag-stained neurons (black color), TH-stained neurons (brown color), or TH/A-Cu-Ag double-stained neurons (silver-stained cell body surrounded by TH staining).

4

Notes 1. Always use glass materials; avoid the use of plastic because silver can strongly bind to plastic [7]. 2. Use high quality water (H2Obd or H2Od) for all steps [7]. 3. Never wash materials with soap or detergent because they can interact with silver and affect the staining quality. Wash glassware with a solution of 30% nitric acid in H2Od [7]. 4. Never use sodium hydroxide (NaOH) or hydrochloric acid (HCl) to adjust the pH of the solutions; sodium and chlorine ions can react with silver ions and interfere with the staining. 5. Use a fresh fixation solution for each experiment, prepared a few hours before the animals will be sacrificed [6]. 6. Instead of cutting brains in coronal sections at 30–35 μm of thickness (as described in de Olmos et al. 1994), cut it in at 50 μm of thickness, to avoid that the slices break during tissue processing. 7. Pay attention to the post-fixation time of the slices in 4% PFA; less than 24 h can enhance the staining of non-degenerating elements, but longer than 24 h can reduce the silver signal in degenerating elements [7]. 8. For an optimal silver precipitation, prepare the pre-impregnation solution 1 day before use and store it at 4  C [7]. 9. 100 mL of pre-impregnation solution should stain around 50 brain slices [7]. 10. Incubation time in the impregnation solution is important; shorter times increase the staining of normal neuronal elements while longer times reduce the staining in degenerating areas [14]. Incubate in the impregnation solution in the dark, as the silver signal decreases with exposure to light. 11. During incubation in the reduction solution, add the impregnation solution (0.3 mL per 100 mL reduction solution) every 5 min to help the silver fix [7].

Amino Cupric Silver (A-Cu-Ag) Staining Neurotoxicity

17

12. Perform the bleaching process 1 day after the reduction reaction [7]. 13. Use oxalic acid to bleach the porcelain after the potassium permanganate has passed through it. 14. The magnification lens to use can vary depending on the size of the target structure. The striatum is a big nucleus that can be analyzed with the 4 objective. Smaller structures need bigger magnification. 15. This stereological estimation is an unbiased method not affected by the volume of the structure or the size of the elements [24]. Is important that the researcher is blind to experimental group conditions, in order to avoid subjective analyses. 16. Stereological counts in the SNpc are done in one hemisphere of the brain, and the data are multiplied by two. This approach can only be done in bilateral structures, if the region of interest is not bilateral, cells must be count in the entire nucleus.

Acknowledgments This protocol was set up in the laboratory of Dr. Rosario Moratalla at Cajal Institute (CSIC) from the protocol described by de Olmos and colleagues in 1994 [14]. We thank Manuel Marquez-Rivera for his help with the revision of the manuscript. Preparation of this manuscript was supported by grants from the Spanish Ministries of Innovation, Science and Universities PID2019-111693RB-I00 and PCIN-2015-098 and Health, Social Services and Equality (PNSD 2016/033 and CIBERNED CB06/05/0055) and UE (H2020-SC1-BHC-2018-2020, grant agreement n 848002). References 1. Switzer RC (2000) Application of silver degeneration stains for neurotoxicity testing. Toxicol Pathol 28(1):70–83 2. Fearnley JM, Lees AJ (1991) Ageing and Parkinson’s disease: substantianigra regional selectivity. Brain 114(Pt 5):2283–2301 3. Beltramino CA, de Olmos JS, Gallyas F et al (1993) Silver staining as a tool for neurotoxic assessment. NIDA Res Monogr 26:136–101; discussion 126–32 4. Bielschowsky M (1904) Silber impregnation der neurofibrillen. J Psychol Neurol 3:169–188 5. Nauta W, Gygax PA (1951) Silver impregnation of degenerating axon terminals in the

central nervous system (1) technic (2) chemical notes. Stain Technol 26:5–11 6. de Olmos JS (1969) A cupric-silver method for impregnation of terminal axon degeneration and its further use in staining granular argyrophilic neurons. Brain Behav Evol 2:213–237 7. de Olmos JD, Beltramino CA, de Olmos-de Lorenzo S (1994) Use of an amino-cupric-silver technique for the detection of early and semiacute neuronal degeneration caused by neurotoxicants, hypoxia, and physical trauma. Neurotoxicol Teratol 16:545–561 8. Eiland MM, Ramanathan L, Gulyani S et al (2002) Increases in amino-cupric-silver

18

Rosario Moratalla et al.

staining of the supraoptic nucleus after sleep deprivation. Brain Res 945(1):1–8 9. Breslow E (1973) Metal-protein complexes. In: Eichhorn GL (ed) Inorganic biochemistry. Elsevier, Amsterdam, The Netherlands, pp 227–249 10. Freeman HC (1973) Metal complexes of amino acid and peptides. In: Eichhorn GL (ed) Inorganic biochemistry. Elsevier, Amsterdam, The Netherlands, pp 121–166 11. Leigh GJ (1990) Nomenclature of inorganic chemistry (recommendations 1990)-the red book. Blackwell Science, London 12. Gallyas F (1982) Physico-chemical mechanism of the argyrophil III reaction. Histochemistry 74(3):409–421 13. Ares-Santos S, Granado N, Espadas I, Martinez-Murillo R, Moratalla R (2014) Methamphetamine causes degeneration of dopamine cell bodies and terminals of the nigrostriatal pathway evidenced by silver staining. Neuropsychopharmacology 39 (5):1066–1080 14. Ramo´n y Cajal S, de Castro F (1933) Elemento de te´cnica microgra´fica del sistema nervioso. Tipografı´a Artı´stica, Madrid 15. de Olmos JS, Ingram WR (1971) An improved cupric-silver method for impregnation of axonal and terminal degeneration. Brain Res 33:523–529 16. Tenkova TI, Goldberg MP (2007) A modified silver technique (de Olmos stain) for assessment of neuronal and axonal degeneration. Methods Mol Biol 399:31–39 17. de Olmos S, Bender C, de Olmos JS, Lorenzo A (2009) Neurodegeneration and prolonged immediate early gene expression throughout cortical areas of the rat brain following acute administration of dizocilpine. Neuroscience 164(3):1347–1359 18. Carmena A, Granado N, Ares-Santos S, Alberquilla S, Tizabi Y, Moratalla R (2015) Methamphetamine-induced toxicity in indusium griseum of mice is associated with astroand microgliosis. Neurotox Res 27 (3):209–216 19. Mendieta L, Granado N, Aguilera J, Tizabi Y, Moratalla R (2016) Fragment C domain of tetanus toxin mitigates methamphetamine neurotoxicity and its motor consequences in mice. Int J Neuropsychopharmacol 19(8): pyw021

20. Granado N, Ares-Santos S, Tizabi Y, Moratalla R (2018) Striatal reinnervation process after acute methamphetamine-induced dopaminergic degeneration in mice. Neurotox Res 34 (3):627–639 21. Granado N, Ares-Santos S, O’Shea E, VicarioAbejo´n C, Colado MI, Moratalla R (2010) Selective vulnerability in striosomes and in the nigrostriatal dopaminergic pathway after methamphetamine administration : early loss of TH in striosomes after methamphetamine. Neurotox Res 18(1):48–58 22. Fujikawa DG, Zhao S, Ke X, Shinmei SS, Allen SG (2010) Mild as well as severe insults produce necrotic, not apoptotic, cells: evidence from 60-min seizures. Neurosci Lett 469:333–337 23. Ares-Santos S, Granado N, Oliva I, O’Shea E, Martin ED, Colado MI, Moratalla R (2012) Dopamine D(1) receptor deletion strongly reduces neurotoxic effects of methamphetamine. Neurobiol Dis 45(2):810–820 24. Granado N, O’Shea E, Bove J, Vila M, Colado MI, Moratalla R (2008) Persistent MDMAinduced dopaminergic neurotoxicity in the striatum and substantia nigra of mice. J Neurochem 107(4):1102–1112 25. Granado N, Ares-Santos S, Oliva I, O’Shea E, Martin ED, Colado MI, Moratalla R (2011) Dopamine D2-receptor knockout mice are protected against dopaminergic neurotoxicity induced by methamphetamine or MDMA. Neurobiol Dis 42(3):391–403 26. Zamanian JL, Xu L, Foo LC et al (2012) Genomic analysis of reactive astrogliosis. J Neurosci 32:6391–6410 27. O’Callaghan JP, Kelly KA, VanGilder RL et al (2014) Early activation of STAT3 regulates reactive astrogliosis induced by diverse forms of neurotoxicity. PLoS One 9(7):e102003 28. Bender C, de Olmos S, Bueno A, de Olmos J, Lorenzo A (2010) Comparative analyses of the neurodegeneration induced by the non-competitive NMDA-receptor-antagonist drug MK801 in mice and rats. Neurotoxicol Teratol 32(5):542–550 29. Sigwald EL, Bignante EA, de Olmos S, Lorenzo A (2019) Fear-context association during memory retrieval requires input from granular to dysgranular retrosplenial cortex. Neurobiol Learn Mem 163:107036

Amino Cupric Silver (A-Cu-Ag) Staining Neurotoxicity 30. Ferna´ndez MS, de Olmos S, Nizhnikov ME, Pautassi RM (2019) Restraint stress exacerbates cell degeneration induced by acute binge ethanol in the adolescent, but not in the adult or middle-aged, brain. Behav Brain Res 364:317–327 31. Rivarola ME, de Olmos S, Albrieu-Llina´s G et al (2018) Neuronal degeneration in mice induced by an epidemic strain of Saint Louis encephalitis virus isolated in Argentina. Front Microbiol 7:9,1181 32. Darmopil S, Martı´n AB, De Diego IR, Ares S, Moratalla R (2009) Genetic inactivation of

19

dopamine D1 but not D2 receptors inhibits L-DOPA-induced dyskinesia and histone activation. Biol Psychiatry 66(6):603–613 ˜ o-Vera E, 33. Espadas I, Darmopil S, Vergan Ortiz O, Oliva I, Vicario-Abejo´n C, Martı´n ED, Moratalla R (2012) L-DOPA-induced increase in TH-immunoreactive striatal neurons in parkinsonian mice: insights into regulation and function. Neurobiol Dis 48 (3):271–281 34. Albrecht M (1954) Mounting frozen sections with gelatin. Stain Technol 29:89–90

Chapter 2 Assessment of Auditory Hair Cell Loss by Cytocochleograms Aure´lie Thomas, Thomas Venet, and Benoıˆt Pouyatos Abstract The highly differentiated cells that compose the sensory epithelium of the cochlea, known as hair cells, are a wonder of refinement, but they are also highly vulnerable because of their location in the cochlea. These cells are indeed subjected both to mechanical stress, through the vibration of the basilar membrane at the base of sound transduction, and to chemical exposure, through the imperfectly protective blood–labyrinth barrier of the stria vascularis. Cochlear histology is therefore a critical step in the assessment of the consequences of ototoxic insults or noise overexposure in animal models. As opposed to other sensory systems, the cochlear neuroepithelium, the organ of Corti, can be harvested with simple tools and every single sensory cell can be observed on a microscope slide, thanks to their regular arrangement. Counting missing hair cells and locating them on the place-frequency map of the considered species provides a wealth of information, both quantitative and objective, on the type and severity of the exposure. This article describes a method to (1) harvest the cochlea, (2) stain and dissect the organ of Corti, (2) quantify the hair cell loss, and (3) construct a graphic representation of the distribution of hair cell loss along the length of the organ of Corti, the cytocochleogram. Key words Cytocochleogram, Method, Dissection, Hearing loss, Cochlea, Organ of Corti, Hair cell count

1

Introduction The hearing system of mammals is composed of a peripheral receptor, the cochlea, which transduces sound waves into neurochemical influxes, neuronal fibers that convey the information towards the brain, where dedicated structures treat the incoming electrical signal and extract meaningful information from it. Although impairment can occur at different levels within this functional chain, the cochlea is the structure most sensitive to environmental stressors because (1) it is directly exposed to mechanical stimuli and because (2) cochlear cells are limited in number and cannot regrow (at least in mammals). Therefore, to investigate the cause of hearing

Supplementary material: The online version of this chapter (https://doi.org/10.1007/978-1-0716-1637-6_2) contains supplementary material, which is available to authorized users. Jordi Llorens and Marta Barenys (eds.), Experimental Neurotoxicology Methods, Neuromethods, vol. 172, https://doi.org/10.1007/978-1-0716-1637-6_2, © Springer Science+Business Media, LLC, part of Springer Nature 2021

21

22

Aure´lie Thomas et al.

loss and/or the consequences of noise and/or chemical exposures, one must start by assessing cochlear health. This can be done by functional or histological methods, these two approaches being, of course, complementary. Functional assessment methods include Distortion Product Otoacoustic Emissions (DPOAE) [1], Auditory Brainstem Responses (ABR) [2] and reflex modification of the startle response [3], but only DPOAEs originate specifically from the cochlea, more precisely from outer hair cells (OHCs). The histology of the cochlea can be carried out by two main techniques: the cytocochleogram, which consists in counting the hair cells on a dissected organ of Corti placed between slide and coverslip [4], and the otic microscopy, which involves embedding the whole cochlea in a solid medium (resin or paraffin and cutting it in thin slices observable with a light or electron microscope. Another alternative is the block surface technique, which involves the slicing of plastic(Araldite-) embedded whole-mount cochlea for observation with a phase-contrast microscope [5]. This method uses the perfusion of the perilymphatic spaces of the cochlea in vivo and avoids postmortem autolytic changes. It is recommended for the assessment of subtle subcellular impairment. If your goal is to obtain an objective and quantitative assessment of hair cell loss (cell present or absent), then the cytocochleogram method described in this chapter is more adapted as it requires minimal equipment and is relatively simple, provided suitable dissecting skills. The method described below can be applied in most laboratory rodents, with varying degrees of difficulty, the cochleae of rats and guinea pigs being easier to process than that of mice for example. The cytocochleogram method allows the whole cochlea to be explored, as opposed to functional methods, which are commonly limited in frequency (the lowest and highest frequencies being, in most cases, inaccessible). Finally, the two ears constitute, in general, two identical replicates, which can save an experiment in case of the unsuccessful dissection of one of the cochleae. There are also some inherent limitations to this method that should be taken into account before undertaking cytocochleograms: this technique is designed to count hair cells and does not allow the assessment of the state of stereocilia or subcellular structures. Also, one should wait a few weeks after exposure or treatment to be certain that all impaired cells have died and have been cleaned out from the organ of Corti before sacrificing the animals and undertaking cytocochleograms (see Note 1). Although the dissection of the unstained organ of Corti is technically possible [6], it is much easier with the use of a cellular staining. The original cytocochleogram method involved the use of post-fixation with osmic acid [7], which turns the cellular lipids black by creating –C¼C– double bonds in the unsaturated fatty acids. While osmic acid creates sufficient contrast to allow the dissection, and is compatible with intracardiac perfusion, it is also

How to make a Cytocochleogram?

23

extremely toxic for the experimenter. Consequently, we favor an alternative method: the succinate dehydrogenase (SDH) staining [8]. The main advantage of this staining is that it the hair cells have a much stronger SDH activity than the surrounding cells (outer sulcus and stria vascularis), which facilitates grandly dissection and counting, given that they are the only “blue” cells in the organ of Corti. In addition, SDH staining does not render the organ of Corti brittle as the osmic acid staining does. The downsides are that the staining has to be immediately done right after cochlear harvest and that the 1-h incubation at 37  C leaves time for the hair cell morphology to potentially evolve. Once the organ of Corti has been harvested from the cochlea and the counting has begun, it is important to be able to place the missing cells in the cochlear place-frequency map (or tonotopic) map of the species considered. The exact knowledge of the cochlear place-frequency map is indeed a prerequisite for the interpretation of normal and abnormal structural and functional features of the inner ear. Such maps have been established in a number of species: rat [9], mouse [10], guinea pig [11], mole rat [12], cat [13], mustache bat [14], fat tailed gerbil [15], opossum [16], Mongolian gerbil [17]. In this chapter, we will focus on the rat, for which the mathematical function describing the place-frequency data is x ¼ 102:048  e ð0,04357f Þ  4:632 where x is the characteristic place expressed in percentage of the basilar membrane length from the base, and f is in the characteristic frequency in kHz [9]. Given the biological variation in the basilar membrane length within the same strain, it is important to standardize the length of the individual cytocochleogram (see Subheading 4.3). First, cochlear length has to be converted to percent distance from base or apex, and then averaged. Cytocochleograms should display both the distance and the characteristic frequency scales. The mathematical function describing the place-frequency data should be indicated in the text or the source given in the reference list. The method described below works well in our laboratory, but might be adapted to the scientific context, (species, strain, age, type of exposure. . .), the available equipment and personnel in the reader’s lab. We provide a 20 min accompanying video showing the dissection of a rat cochlea. It can be found there: https://youtu.be/ QaDFwHjYTbQ. Other videos showing the dissection of mouse cochleae are also available on the internet [6, 18].

24

2

Aure´lie Thomas et al.

Materials l

Equipment and tools: – laboratory balance (Mettler Toledo XP205). – pH-meter with 0.01 pH precision (Labo Moderne). – dentist’s drill (NSK Volvere max NE120) equipped with diamond heads (Ø1.4 and 0.6 mm). – heating water bath. – light microscope equipped with a micrometric eyepiece (Olympus BX41). – calibration slide (2.5 μm/graduation). – friedman rongeur. – cutting edge. – bone scissors. – 23G needles. – fine forceps.

l

Chemicals: – Nitro blue tetrazolium chloride (NBT) (Merck, Cat #1.24823). – Sodium succinate (Merck, Cat #8.20151). – Potassium phosphate monobasic (KH2PO4) (Merck, Cat #1.04873). – Dipotassium hydrogenphosphate (K2HPO4) (Sigma, Cat #60353). – Phosphate-buffered saline (PBS). – Paraformaldehyde 32% (EMS, Cat #15714-S). – Decalcifying solution (Cellpath, Cat #DC-LMR) (optional).

3

Sample Preparation

3.1 Preparing Solutions

Prepare the following stocks solutions and store at 4  C: – 0.1% Nitro blue tetrazolium chloride (NBT): 0.5 g for 500 mL of distilled water. – 0.2 M sodium succinate: 27.015 g for 500 mL of distilled water. – 0.2 M KH2PO4 solution: 13.6 g of KH2PO4 for 500 mL of distilled water. – 0.2 M K2HPO4 solution: 34.84 g of K2HPO4 for 1000 mL of distilled water.

How to make a Cytocochleogram?

25

Mix 190 mL of KH2PO4 solution and 810 mL of K2HPO4 solution to obtain the final 0.2 M phosphate buffer, and adjust at pH 7.4. The day of the sacrifice, prepare a solution of 4% paraformaldehyde in 1 PBS and the SDH staining solution (0.1% NBT, 0.2 M sodium succinate, and 0.2 M phosphate buffer with a 2:1:1 ratio). Ensure the SDH solution has an osmolality of about 280 mmol/ kg. Approximately 100 mL of staining solution is necessary for two rat cochleae. 3.2

Cochlea Harvest

Sacrifice the animal according to local ethical regulations for animal experiments (i.e., EU Directive 2010/63/EC). Cut the skull in the longitudinal axis with strong bone scissors and remove the brain (Fig. 1a). Locate the external auditory canal and cut the temporal bone around the cochlea with a large safety margin (Fig. 1b). Repeat for the other ear. Remove the masseter muscle by pulling the flesh with the rongeur (Fig. 1c) to reveal the otic capsule (Fig. 1d). Carefully break the capsule using the rongeur. Once the cochlea is visible in the open otic capsule, continue to clear the space around it by carefully removing the stapes, the basilar artery, and bone debris to access round and oval windows (Fig. 1f). Be sure you have a good hold on the temporal bone between your thumb and index finger to be comfortable for the dissection to come. To do so, remove as much flesh from the bone as you can.

3.3

SDH Staining

Immediately after cochlea collection, immerse the samples in the SDH staining solution. Access to both the round and oval windows must be clear. Gently perfuse both windows with the SDH staining

Fig. 1 Different steps of cochlea harvest

26

Aure´lie Thomas et al.

Fig. 2 Light microscope views of a cochlea at different steps of dissection. (a) Cochlea before drilling. (b) Cochlea drilled. (c) Cochlea without bone. (d) Cochlea without stria vascularis, Reissner, and tectorial membranes. Hair cells stained in dark blue are clearly visible

solution using a syringe and needle with a smoothed tip. Some blood usually escapes from both windows. This ensures that the cochlea is properly perfused. Then, immerse the cochleae in the same solution for 1 h at 37  C. After staining, fix the cochleae by injecting 4% paraformaldehyde in both windows and keep 24 h at +4  C. Finally, rinse with PBS and store at +4  C until microdissection. 3.4 Cochlear Microdissection

4

Place the temporal bone in a large cup containing 1 PBS under a light microscope (Fig. 2a). Before microdissection, carefully drill the cochlea using a dentist’s drill equipped with a diamond head to leave only a thin layer of bone (see Fig. 2b, Note 2 and video https://isicloud.inrs.fr/index.php/s/m57h4j4bC2QcdkJ). Gently drill a small hole at the top (apex) of the cochlea. Carefully remove tiny chunks of bone using a cutting-edge starting from the apical hole and progressing slowly towards the base. Hair cells should be clearly visible with their dark blue color when the organ of Corti is exposed (Fig. 2c). With the tip of a needle (23G), cut the stria vascularis at the level of the outer sulcus. The Reissner and tectorial membranes, completely transparent with this staining, have to be removed entirely with fine forceps to facilitate subsequent counting (Fig. 2d). Using the 23G needle cut the organ of Corti and separate it from the osseous spiral lamina with a fine cutting edge (see Note 3). Then cut the organ of Corti in three parts (apical, medial, and basal turns) before immersing them in PBS. Trim the bone as close to the organ of Corti as you safely can and remove all remaining pieces to stria vascularis and other membranes. Now put the sample on a glass slide in glycerin and place a cover glass on top (Fig. 3). The three turns can be mounted on a single slide.

Data Collection and Analysis The creation of a cytocochleogram involves several successive steps:

How to make a Cytocochleogram?

27

Fig. 3 The three turns of the rat cochlea after microdissection. (a) Apical turn, (b) Medial turn, and (c) Basal turn

1. Calibration. 2. Hair cell count (missing and/or present). 3. Normalization of the cochlear length (for averaging). 4. Calculation of percent loss of hair cells per unit of distance. 5. Plotting. These steps require either a standardized Excel file or a dedicated computer interface (Fig. 4), with which you can enter the positions of the missing hair cells along the cochlea (see Note 4). 4.1 Microscope and Software Calibration

Place the calibration slide under the light microscope equipped with a micrometric eyepiece and a 40 objective. Record the distance covered by the 100 graduations. This step needs to be performed every time a new microscope or a new objective lens is used. Now, place the surface preparation under the microscope. The first step is to determine how many outer and inner hair cells are visible within the 100 graduations of the eyepiece micrometric scale. Because of the difference in hair cell size between the apex

28

Aure´lie Thomas et al.

Fig. 4 (a) Microscopic view of dissected organ of Corti with the superimposed eyepiece micrometric scale places on top of the outer hair cell rows. For the two inner rows of outer hair cells (OHC1 and OHC2), missing hair cells appear in black on the software interface (b). For the third row of outer hair cells, extremely damaged, the remaining outer hair cells are counted and appear in white on the software (OHC3). No missing inner hair cell (IHC) is visible. In this case, the button “All present” is pressed. The data input can performed in simple Excel sheet (c). This example corresponds to the same cochlea partition visible on the photograph and analyzed in the software interface

Table 1 Number of hair cells for 100 graduations in the three turns of the cochleae of Brown-Norway rats (6–24 months old). Data are mean (n ¼ 82)  standard error Inner hair cells

Outer hair cells

Apical turn

28.6  1.23

37.1  1.63

Medium turn

31.0  1.46

37.7  1.39

Basal turn

31.1  1.63

37.2  1.65

How to make a Cytocochleogram?

29

and the base and between inner and outer hair cells, record six values, one for each cochlear turn and for each type of hair cell. For example, the average values obtained in adult Brown-Norway rats are depicted in Table 1. At this point, you dispose, for each turn, of a correspondence table between the number of graduations, the number of hair cells, and the real distance on the organ of Corti. Before starting to count the missing hair cells, you should record the correspondence table in the software interface/Excel Sheet, as well as the information concerning your sample (species/ strain, left or right cochlea, reference number/code of the subject, treatment). 4.2

Hair Cell Count

Once the calibration is performed, counting can start. The aim of this step is to obtain the location of each missing cell, expressed as an absolute distance from the apical end of the organ of Corti. To do so, place the graduation at the extreme apex of the cochlea, zero on the first visible outer hair cell (see Note 5). This position will be considered as the “position zero,” and the positions of all missing cells will be expressed as an absolute distance from this “position zero.” The apical turn of the organ of Corti being more curved than the other turns, you must count over a distance shorter than the full scale of the eyepiece, e.g., 50 graduations. In any case, the number of graduations should be registered (Fig. 4c; column #2) and the starting / ending graduations of the section should be converted as a distance from the “position zero” (Fig. 4c; columns #3 and #4) using your correspondence table (in our case: 100 graduations ¼ 250 μm). The locations of the missing hair cells should always be recorded in order (from apex to base). For each hair cell row (column #5), locate the starting graduation of each “hole” (Fig. 4c; column #6). It is then convenient to be able to record either the number of contiguous missing hair cells (for small “holes”; column #8) or the graduations of the end of the series of missing hair cells (for big “holes” for which the number of missing hair cell is difficult to know precisely; column #7). In either case, the software (or the Excel sheet) should convert the starting and ending graduations into a position relative to the “position zero” (column #10 and #11). In case of an intact hair cell row (Fig. 4a; inner hair cell) or an important number of missing cells (Fig. 4a; third row of outer hair cells), it is much easier and faster to count the remaining hair cells instead of the missing ones (Fig. 4b,

30

Aure´lie Thomas et al.

OHC3; Fig. 4c, column #9). When the counting is finished for the section considered, save the data and progress to the next section (see Note 5). 4.3 Cytocochleogram Construction and Display

Once the number of missing (or present) hair cells and their locations are registered in the computer interface of your choice, it is time to construct the cytocochleogram. To do so, be sure to dispose of the following data: – the total length of the cochlea. – the reference length for an average cochlea (e.g., 9.3 mm for a rat and 18.5 mm for a guinea pig). – the starting and ending positions of “holes” relative to “position zero”. – the mathematical function of the place-frequency map of the species considered. – the number of cells per 100 graduations for each turn. The cytocochleogram is generally represented as a bar chart or a line plot, with the percentage of hair cell loss in ordinate and the position along the length of the cochlea in abscissa. The abscissa should also include a second x-axis displaying the frequency calculated using the place-frequency map equation. Therefore, the length of the cochlea has to be divided into “sections,” which will be represented by each bar/point of the graph. The size of each section depends on the precision needed to depict the cochlear damage accurately. A default size of 100 μm is adequate in most cases. To calculate the data to be plotted in the cytocochleogram, go through the following steps: 1. Read the positions of the “holes” in the different cell rows. In case the positions of “present cells” have been recorded, consider the intervals to get the positions of the “holes”. 2. Normalize the positions of the “holes” against the reference length of the cochlea using a linear interpolation. Let us consider a series of missing cell between 7000 and 7050 μm on a cochlea measuring 9.6 mm. The normalized positions will be. 7000  9:3=9:6 ¼ 6781:25 μm and 7050  9:3=9:6 ¼ 6829:69 μm:

3. Affect each “hole” to the sections to be displayed in the cytocochleogram (Fig. 5). (a) Case a: If one or several “holes” are entirely contained within a section, the percentage of missing cells for this

How to make a Cytocochleogram?

31

100 graduations

Hair cell loss [%]

HAIR CELLS

HAIR CELLS

100 90 80 70 60 50 40 30 20 10 0

0

100

200

300

400

500

600

700

800

900

1000 1100

1200

1300

1400 1500

1600

1700

Position on the organ of Corti [mm]

Fig. 5 Assignment of missing cells to the sections of a cytocochleogram. Red hair cells illustrate the missing cells. The cytocochleogram is plotted with 100 μm sections

section is directly calculated. Let us consider the section 1500–1600 μm (Fig. 5). The percentage of missing cells is [(1664–1630) + (1695–1679)]/100 ¼ 50%. (b) Case b: If the two extremities of the “hole” are not contained within a single section, distribute the distance in the sections overlapped by the “hole” and calculate the percentage of missing cells. Let us consider the “hole” overlapping sections 600–700, 700–800, and 800–900 μm in Fig. 5. The percentage of missing cells in section 600–700 μm is [(618–600) + (700–693)]/ 100 ¼ 25%. One hundred percent of the cells are missing in section 700–800 μm. Then, the percentage of missing cells in section 800–900 μm is (916–900)/100 ¼ 16%. 4. Use the place-frequency mathematical equation to convert the distance into frequencies and create a second x-axis (kHz). 5. Plot the different rows of hair cells in four different subplots (Fig. 6). Length standardization allows the calculation of average cytocochleograms and the application of statistical analyses. Cytocochleograms can be plotted with error bars.

5

Conclusions There is no doubt that creating cytocochleograms takes a bit of effort. The drilling and the dissection of the cochlea require some training. The programming of the counting interface may also be intimidating. However, there is no unachievable step in this method, and it can be performed with the standard laboratory equipment and chemicals. The cochlea is arguably the only sensory organ of the body, whose histological status can be assessed quantitatively, thanks to the quasi-linear organization of the organ of Corti. Although the cytocochleograms only provide information on the missing hair

32

Aure´lie Thomas et al.

Fig. 6 Example of an average (n ¼ 5) cytocochleogram obtained in male Brown-Norway rats exposed to 600-ppm styrene and an impulse noise with a LEX, 8 h (equivalent continuous noise level averaged over 8 h) of 80 dB for 4 weeks, 5 days per week, 6 h per day by inhalation. Rats were sacrificed 4 weeks after the end of exposure. Error bars represent the standard error mean

cells and their location, and do not give any insight on the ganglion cells, it is possible (and wise) to use the contralateral ear to perform otic microscopy to observe there any additional neuronal damage. Still, there are a few constraints and limitations inherent to the cytocochleogram. In case the goal is to assess cochlear damage caused by either a chemical treatment or noise, consider waiting at least 4 weeks after the end of the exposure to harvest the cochlea. Logically, the cytocochleogram only allows the assessment of the permanent consequences of an exposure, and the exposure has to be severe enough to cause hair cell loss. If you are interested subtle temporary auditory effects, you should turn to functional methods (otoacoustic emissions, brainstem auditory evoked potential) which give you the possibility to perform repeated measures at different time points. Ultimately, the mastery of cytocochleograms is a great benefit for a hearing research lab and only requires a bit of initial training and set up.

How to make a Cytocochleogram?

6

33

Notes 1. This process takes at least 4 weeks, but we recommend waiting a few weeks longer, if it is experimentally possible. 2. Take all the time needed for this step because the finer the bone layer is, the easier the dissection will be. It is better to make few holes with the drill in the bony shell, than leaving too much bone. 3. Please note that the “hook” of the basal turn of the organ of Corti is more difficult to dissect than the rest of the organ because the spiral lamina is thicker. You might want to soften the bone by immersing the cochlea for 5–10 min in the decalcifying solution before cutting the spiral lamina with the 23G needle. In case you are having difficulties with this dissection, it is also possible to cut the basal turn into reasonable segments with iris scissors to avoid tearing the organ of Corti although this method will cause more tissue loss than the dissection in one single piece. 4. If you plan to perform cytocochleograms on a regular basis, automatizing the different steps using a little programming (e.g., VBA Excel; Fig. 4b), is very helpful to save time and avoid mistakes. However, it also possible to use a simple Excel file, and manually apply the different calculations needed to obtain the data for plotting the cytocochleogram (Fig. 4c). 5. The hair cells of the extreme apical and basal ends of the cochlea appear often disorganized. In the rat, the apical outer hair cells are often widely spaced and unaligned. At the basal end, the third row of outer hair cells often stops before the second and first rows. Consider these features as normal. Note that these cellular disarrays worsen as the animals get older and can represent the first signs of presbycusis.

References 1. Venet T, Thomas A, Wathier L, Pouyatos B (2020) DPOAEs for the assessment of toxicant-induced cochlear damage in Neuromethods, Vol. 172, Jordi Llorens and Marta Barenys (Eds): Experimental Neurotoxicology Methods, ISBN 978-1-0716-1636-9 (Chapter 5) 2. Eggermont JJ (2019) Auditory brainstem response. Handb Clin Neurol 160:451–464. https://doi.org/10.1016/B978-0-44464032-1.00030-8 3. Davis M (1984) The mammalian startle. In: Eaton RC (ed) Neural mechanisms of startle behavior. Springer, Boston, MA, p 287

4. Engstro¨m H, Ades HW, Andersson A (1966) Structural pattern of the organ of Corti: a systematic mapping of sensory cells and neural elements. Almqvist & Wiksell, Stockholm 5. Bohne BA (1986) The plastic-embedding technique for preparing the chinchilla cochlea for examination by phase-contrast microscopy. Washington University Laboratory Manual vol. 6; https://www.researchgate.net/profile/ Barbara_Bohne/publication/237197190_THE_ PLASTIC-EMBEDDING_TECHNIQUE_ FOR_PREPARING_THE_CHINCHILLA_ COCHLEA_FOR_EXAMINATION_BY_ PHASE-CONTRAST_MICROSCOPY/

34

Aure´lie Thomas et al.

links/5755394308ae0405a5736685/ THE-PLASTIC-EMBEDDING-TECH NIQUE-FOR-PREPARING-THE-CHIN CHILLA-COCHLEA-FOR-EXAMINA TION-BY-PHASE-CONTRAST-MICROS COPY.pdf?origin¼publication_detail 6. Liberman MC (2015) Cochlear dissection for whole mount immunostaining. Mass Eye and Ear. https://www.masseyeandear.org/ research/otolaryngology/eaton-peabodylaboratories/histology-core 7. Anniko M, Lundquist PG (1977) The influence of different fixatives and osmolality on the ultrastructure of the cochlear neuroepithelium. Arch Otorhinolaryngol 218(1–2):67–78. https://doi.org/10.1007/bf00469735 8. Yang WP, Hu BH, Sun JH, Zhai SQ, Henderson D (2010) Death mode-dependent reduction in succinate dehydrogenase activity in hair cells of aging rat cochleae. Chin Med J 123 (13):1633–1638 9. Muller M (1991) Frequency representation in the rat cochlea. Hear Res 51(2):247–254. https://doi.org/10.1016/0378-5955(91) 90041-7 10. Muller M, von Hunerbein K, Hoidis S, Smolders JW (2005) A physiological placefrequency map of the cochlea in the CBA/J mouse. Hear Res 202(1–2):63–73. https:// doi.org/10.1016/j.heares.2004.08.011 11. Tsuji J, Liberman MC (1997) Intracellular labeling of auditory nerve fibers in Guinea pig: central and peripheral projections. J Comp Neurol 381(2):188–202

12. Muller M, Laube B, Burda H, Bruns V (1992) Structure and function of the cochlea in the African mole rat (Cryptomys hottentotus): evidence for a low frequency acoustic fovea. J Comp Physiol A 171(4):469–476. https:// doi.org/10.1007/bf00194579 13. Liberman MC (1982) The cochlear frequency map for the cat: labeling auditory-nerve fibers of known characteristic frequency. J Acoust Soc Am 72(5):1441–1449. https://doi.org/10. 1121/1.388677 14. Kossl M, Vater M (1985) The cochlear frequency map of the mustache bat, Pteronotus parnellii. J Comp Physiol A 157(5):687–697. https://doi.org/10.1007/bf01351362 15. Muller M, Ott H, Bruns V (1991) Frequency representation and spiral ganglion cell density in the cochlea of the gerbil Pachyuromys duprasi. Hear Res 56(1–2):191–196. https:// doi.org/10.1016/0378-5955(91)90169-a 16. Muller M, Wess FP, Bruns V (1993) Cochlear place-frequency map in the marsupial Monodelphis domestica. Hear Res 67 (1–2):198–202. https://doi.org/10.1016/ 0378-5955(93)90247-x 17. Muller M (1996) The cochlear place-frequency map of the adult and developing Mongolian gerbil. Hear Res 94(1–2):148–156. https:// doi.org/10.1016/0378-5955(95)00230-8 18. Fang Q-J, Wu F, Chai R, Sha S-H (2019) Cochlear surface preparation in the adult mouse. JoVE. https://doi.org/10.3791/ 60299. https://www.jove.com/video/60299

Chapter 3 Evaluation of Cellular and Molecular Pathology in the Rodent Vestibular Sensory Epithelia by Immunofluorescent Staining and Confocal Microscopy Alberto F. Maroto, Erin A. Greguske, Alejandro Barrallo-Gimeno, and Jordi Llorens Abstract Hair cells in the vestibular and cochlear sensory epithelia are the main target of ototoxic drugs. Nevertheless, the synapses between the hair cells and the afferent terminals of the post-synaptic ganglion neurons have also been shown to be a target of ototoxic damage. In this chapter, we describe immunohistochemistry protocols adapted to the quantification of hair cells and synapses in the vestibular epithelia to assess ototoxic damage in rodents. Epithelia are immunolabeled intact and are used in whole-mount preparations for the quantification of hair cell numbers by confocal microscopy imaging. For synaptic assessment, the epithelia are first immunolabeled, embedded in a gelatin/albumin block, and then sectioned in a vibrating microtome before confocal microscopy imaging. The data thus obtained offer a thorough evaluation of the damage suffered by the vestibular sensory epithelia, including overt hair cell loss and subtle synaptic loss. Together, these pathological outcomes determine the loss vestibular input and the resulting behavioral alterations. Key words Ototoxicity, Immunohistochemistry, Confocal microscopy, Vestibular epithelia, Hair cell, Synaptic uncoupling

1

Introduction

1.1 Hair Cell and Synapse Counts in the Vestibular Sensory Epithelia Following Ototoxic Damage

The vestibular system in the inner ear, also known as labyrinth, detects head accelerations resulting from body movements and gravity [1]. It provides both dynamic and static information for equilibrium and gaze control, as well as for the sense of orientation in space. Vestibular dysfunctions cause disequilibrium and loss of gaze control and are accompanied by vertigo, dizziness, impaired visual acuity, compromised motor competence, falls, and other disabilities. One possible cause of vestibular loss is exposure to toxic chemicals that target both the vestibular and the auditory systems, hence named ototoxic compounds [2]. Besides their shared location in the inner ear, the vestibular and the auditory

Jordi Llorens and Marta Barenys (eds.), Experimental Neurotoxicology Methods, Neuromethods, vol. 172, https://doi.org/10.1007/978-1-0716-1637-6_3, © Springer Science+Business Media, LLC, part of Springer Nature 2021

35

36

Alberto F. Maroto et al.

systems rely for sensory transduction on subtypes of a common cellular type, the hair cell (HC). HCs are specialized epithelial cells characterized by bundles of apical specializations, known as stereocilia, that protrude into fluid-filled cavities and contain the molecular machinery for mechanoelectrical transduction [3]. Well known ototoxic compounds include clinically important drugs, such as aminoglycoside antibiotics and the chemotherapeutic drug cisplatin. Several different aminoglycosides have clinical use and all of them are ototoxic although some are more toxic to the auditory system and others, notably gentamicin and streptomycin, affect more the vestibular system. These drugs cause HC degeneration and often permanent loss of function because the ability of HCs to regenerate is null or very limited [4]. In animal models of ototoxicity, counts of sensory HC numbers are obtained for different purposes, such as to evaluate ototoxic potency of drugs, to assess the potential benefit of candidate protective agents that could have clinical utility by reducing the extent of damage, or to measure the efficacy of treatments aimed at activating HC regeneration [5]. In addition, there is growing interest in obtaining counts of synaptic contacts between the HCs and their post-synaptic afferents. Although the HCs are likely the primary target of ototoxic drugs, recent research has revealed that loss of synaptic contacts, referred to as synaptic uncoupling, may account for a significant part of the functional loss in particular ototoxicity models, depending on the ototoxic compound and mode and duration of exposure [6–9]. 1.2 Functional and Structural Diversity in Vestibular Epithelia and Differential Vulnerability to Ototoxic Damage

Two types of HCs are identified in the vestibular sensory epithelia (Fig. 1). Type I HCs (HCI) have an amphora-like shape and are encased up to their neck by a remarkable structure, a calyx-shaped afferent terminal. In the lower two-third of the basolateral membrane of the HCI, the contact between the cell and the inner membrane of the calyx shows an adhesion complex, the calyceal junction, that is characterized at the transmission electron microscopy level as a symmetric densification of the plasma membrane, a greater than normal regularity in the distance between the membranes, and an electro-dense appearance of the extracellular space [10, 11]. Type II HCs (HCII) have a more columnar form and are contacted by bouton afferent terminals that are like regular synaptic boutons found elsewhere in the nervous system. Like HCs in the cochlea and photoreceptors and bipolar cells in the retina, both types of vestibular HCs are associated with a unique structure located in the pre-synaptic active zone, the synaptic ribbon. The ribbon likely contributes to the fast release and/or recycling of the synaptic vesicles thus enabling continuous, robust release of neurotransmitter [12]. The neurotransmitter in these synapses is glutamate, and post-synaptic sites contain ionotropic AMPA-type glutamate receptors clustered in post-synaptic densities [6, 8, 9,

Vestibular Toxicity Assessment by Immunofluorescence

HCII

SC

c

HCII HCI

HCI

HCI

c

b SC

37

SC

Post-synaptic density

b

c

cc

HCI

SC Calyceal junction Calyx-only afferent

b

SC

SC

c

SC

Ribbon Bouton-only afferent

Dimorphic afferent

Fig. 1 Schematic of the various cell types and afferent endings in the vestibular epithelium. The epithelium consists of hair cells (HC), characterized by the stereocilia bundles on their apical surfaces, and supporting cells (SC). Type I HCs (HCI) have an amphora-like shape, are encased by calyceal afferent terminals (c), and show a prominent adhesion complex with this terminal, the calyceal junction. Type II HCs (HCII) are more cylindrical and are contacted by bouton terminals (b). Synaptic contacts are discrete structures, identified by pre-synaptic ribbons and post-synaptic densities. Three types of afferents are found: (1) calyx-only afferents, often forming complex calyces (cc) that encase 2 or 3 HCI; (2) bouton-only afferents, contacting only HCII; and (3) dimorphic afferents that form both calyceal and bouton type terminals

13]. Although there are two major types of HCs and two types of afferent terminals, three classes of afferents can be differentiated [14]. These are: (1) calyx-only afferents that form calyx terminals only and often form complex calyces that engulf two or more HCIs; (2) bouton-only afferents that form bouton terminals on HCII; and (3) dimorphic afferents that form both calyx terminals on HCI and bouton terminals on HCII. Each human or rodent ear contains five vestibular sensory epithelia: three cristae, one in each of three semicircular canals, and two maculae in the utricle and saccule (Fig. 2). While the overall structures of the epithelia are similar in all vestibular end organs, there are regional differences within and among them, especially between the central and peripheral zones of each epithelium [14–17]. The stereocilia bundles are polarized structures, and their orientation reverses at approximately the central line of the utricular macula and the saccular macula, known as the striola. The peri-striolar (central) zone of these maculae is characterized by a higher density of HCI and denser innervation by calyx-only terminals. In the periphery of the maculae, HCII are more abundant and HCI are contacted by calyces that arise from dimorphic afferents. The crista receptors do not have a striola because all HCs in a crista

38

Alberto F. Maroto et al.

Crista

Saccule

Utricle

Crista - Merge

Saccule - Merge

a

b

Utricle - Merge

Utricle – Myo7a

c

d

Utricle - Calretinin

Utricle - Oncomodulin

e

f

Fig. 2 Diversity of the vestibular sensory epithelia within specific end organs. The top drawings are schemas of the crista saccule and utricle, with the central part of the crista and peri-striolar part of the saccule and utricle shown in solid color. The bottom panels show confocal images of these epithelia (a: crista; b: saccule; c–f: utricle) labeled with antibodies against Myo7a (all HCs; red), calretinin (HCII and calyx-only afferents; green) and oncomodulin (HCs in central/peri-striolar zones; blue), as indicated within the figures. The scale bar in a is 100μm and applies to all panels

Vestibular Toxicity Assessment by Immunofluorescence

39

show the same orientation. They do have a central (apical) part with equivalent features to the peri-striolar zone of the maculae, and a periphery similar to the periphery of the maculae. One of the most striking differences between end organs is the length of the stereocilia, which are much longer in cristae than in maculae. These morphological differences associate with still poorly characterized biochemical and physiological differences. One consistent observation across most experimental studies of vestibular toxicity is the differential sensitivity of the different end organs, zones, and cell types to different ototoxic compounds, including aminoglycosides and nitriles [18–21]. In dose–response or time-course studies, damage typically progresses in the following orders: crista > utricle > saccule, central > peripheral and HCI > HCII. Thus, HCI contacted by calyx-only afferents in the apical part of the crista show the highest sensitivity, degenerating at the minimally effective doses. Subsequently, at nearly maximal doses, only a few cells survive the ototoxic treatment, and these are typically HCII in the peripheral zones of the saccule. 1.3 Use of Immunofluorescent Labeling to Assess HC and Synaptic Loss

While scanning electron microscopy observation of the vestibular epithelia surfaces offers a direct overall view of vestibular HC loss [19, 22], immunofluorescent labeling, and subsequent assessment by confocal microscopy together offer the best approach to quantify loss of HCs by type and synaptic damage. This approach relies on knowledge of proteins selectively expressed in each type of HC or in discrete subcellular structures, and the availability of dependable antibodies against these proteins. The most common marker used to selectively label HCs for quantification is the unconventional myosin VIIa (Myo7a) [23]. HC nuclei have also been successfully labeled for HC quantification with antibodies against the transcription factor Gfi1b [5]. Antibodies against other proteins allow differentiation between HC types. HCIs have been shown to selectively express secreted phosphoprotein 1 (SPP1)/osteopontin [17, 24], while most HCII express calretinin [15]. Calyx-only afferents also express calretinin while bouton-only and dimorphic afferents do not [25]. Calyx terminals, and, indirectly HCI, can be visualized by labeling proteins characterizing the calyceal junction, either the adhesion molecule contactin-associated protein 1 (Caspr1), found in the afferent membrane [6, 9, 11, 16], or the extracellular matrix protein, tenascin-C [6, 9, 16]. Synaptic contacts can be recognized as close pairs of pre- and post-synaptic puncta. The core component of the pre-synaptic ribbon is ribeye, a protein encoded by the same gene encoding the transcription factor C-terminal binding protein 2 (CtBP2) [26]. The post-synaptic density is rich in the post-synaptic density protein 95 (PSD-95), a scaffold protein necessary for the clustering of the glutamate AMPA receptors [27]. Shank1A is another protein that characterizes the post-synaptic density [28, 29]. Although

40

Alberto F. Maroto et al.

mRNA expression data [30] suggest that these may contain GluA2, GluA3, and GluA4, but not GluA1 subunits, so far only the GluA2 subunit has been clearly labeled in the vestibular synapses [6, 9]. 1.4 Labeling of the Vestibular Sensory Epithelia

2

Immunofluorescent labeling is a routine procedure in many laboratories, and each laboratory optimizes the basic protocols for their needs, determined by the tissues to be examined and the properties of the antibodies to be used. This chapter will describe the protocols optimized in our laboratory to assess the number of HCs, including separate counts for HCI and HCII, and the number of synaptic contacts. To study the vestibular epithelia, the traditional approach of labeling tissue sections on slides requires these tiny tissues to be embedded in an appropriate medium to form a block for sectioning. Embedding can be done with a medium for cryostat sectioning or with agar for sectioning at the vibrating microtome [31]. The vibratome sections can also be used for free floating labeling [32]. However, the vestibular sensory epithelia are thin enough to allow whole mount immunolabeling and imaging by confocal microscopy through their entire thickness. Therefore, this has become in recent years the most common approach to evaluate HCs and their synaptic densities [7, 8]. Nevertheless, imaging of epithelial cross-sections may be useful for a meticulous investigation of synapses and, for this purpose, we combine whole mount immunolabeling with sectioning of the epithelium labeling [6, 9]. To obtain the sections, we embed the labeled epithelia in gelatin/albumin blocks as often done with zebrafish and Xenopus embryos [33]. The immunolabeling protocol mostly follows that described by Lysakowski et al. in 2011 [16].

Materials

2.1 For Sample Processing and Immunohistochemistry

– Freshly depolymerized paraformaldehyde, 4% in 0.1 M phosphate-buffered saline (PBS), pH 7.2 (see Note 1). – Cryoprotective solution: 34.5% glycerol, 30% ethylene glycol, 20% PBS, 15.5% distilled water. – Triton-X-100 (Sigma-Aldrich). – Gelatin from cold water fish skin, CAS #9000-70-8 (SigmaAldrich). – Mounting medium suitable for fluorescence imaging, such as Mowiol. – Microscope slides and coverslips. – Primary antibodies (Table 1). – Fluorochrome-conjugated secondary antibodies. These can be selected as appropriate for the primary antibodies from a large

Vestibular Toxicity Assessment by Immunofluorescence

41

Table 1 Antibodies Target

Host and type

Reference and source

Working dilution

Calretinin

Guinea pig, polyclonal

214 104, Synaptic Systems

1/500

Calretinin

Rabbit, polyclonal

CR7697, Swant

1/1000

Caspr1

Mouse, monoclonal (IgG1)

Clone K65/35, Neuromab

1/400

GluA2

Mouse, monoclonal (IgG2a)

Clone 6C4, MAB397 Millipore

1/100

Myosin VIIa

Mouse, monoclonal (IgG1)

Clone 138-1-s, DSHB

1/100

Myosin VIIa

Rabbit, polyclonal

25-6790, Proteus Biosciences

1/400

Oncomodulin

Rabbit, polyclonal

OMG4, Swant

1/400

PSD-95

Mouse, monoclonal (IgG2a)

Clone K28/43, Neuromab

1/100

Ribeye

Mouse, monoclonal (IgG1)

Clone 16/CtBP2, BD Biosciences

1/200

SPP1

Goat, polyclonal

AF808, R&D Systems

1/200

Tenascin

Rabbit, polyclonal

AB19013, Millipore

1/200

variety of available choices. For the present protocols, we use the following secondary antibodies conjugated with Alexa Fluor fluorochromes: 488 goat anti-guinea-pig IgG H + L (catalog #A11073, Invitrogen/ThermoFisher), 555 donkey anti-rabbit IgGs H + L (#A-31572), 555 goat anti-mouse IgG2a (#A21137), 647 goat anti-mouse IgG1 (#A21240), and 654 donkey anti-mouse IgG H + L (catalog #A-21202). We also use the DyLight 405 donkey anti-rabbit IgG H + L (catalog #711-475-152, Jackson ImmunoResearch) (see Note 2). – Appropriate nuclear fluorescent stain (i.e., 40 ,6-diamidino-2phenylindole/DAPI). 2.2 For Post-Labeling Sectioning

– Gelatin/albumin mixture for post-labeling inclusion and sectioning of the vestibular epithelia: 0.49 g alimentary gelatin, 30 g bovine serum albumin, 20 g sucrose, 100 mL of 0.1 M PBS (see Note 3). – Glutaraldehyde 25%. – Small molds. We use 1 cm long cylinders obtained by two transverse cuts of 2 mL Eppendorf tubes. – Vibrating microtome.

2.3 For Confocal Imaging and Image Analysis

– Confocal microscope suitable for 4-color channel image acquisition. The images in this chapter were acquired in a Zeiss LSM880 spectral microscope.

42

Alberto F. Maroto et al.

– Microscopy image analysis software. We use both Image J (National Institute of Mental Health, Bethesda, Maryland, USA) and Imaris (Bitplane).

3

Methods

3.1 Immunohistochemistry Protocol

l l

Fix the vestibular epithelia in 4% PFA for 1 h. Rinse the samples in PBS (2  10 min) and use immediately or store at 20  C in cryoprotective solution. For storage, the samples are placed in the solution at room temperature, then placed at 4  C for 2 h for effective embedding before storage in the freezer. To use these samples, allow them to return to room temperature and rinse with PBS (2  10 min). We have successfully used samples stored in this way for up to 8 years.

l

All the following steps are performed using slow agitation.

l

Incubate samples for 1 h with 500μL of 4% Triton X-100 in PBS for permeabilization (see Note 4).

l

Incubate samples for 1 h with 500μL of 0.5% Triton X-100 and 1% of fish gelatin in PBS for blocking (see Note 5).

l

Incubate primary antibodies in 200μL of 0.1% Triton X-100 and 1% fish gelatin in PBS for 24 h at 4  C. Details of the primary antibodies are shown in Table 1. We use two different combinations: – A: To obtain counts of HCI and HCII, we use the rabbit antiMyo7a antibody to label all HCs, the guinea-pig anti-calretinin to label HCIIs and the mouse monoclonal anti-Caspr1 antibody to label calyceal junctions and thus identify HCI. – B: To quantify synaptic contacts, we use the rabbit antiMyo7a antibody to label all HCs, the guinea-pig anti-calretinin to label HCII, the mouse IgG1 anti-ribeye antibody to label pre-synaptic ribbons and the mouse IgG2a anti-PSD95 to label post-synaptic densities.

l l

Rinse with 500μL of PBS (4  10 min). Incubate secondary antibodies in 200μL of 0.1% Triton X-100, 1% fish gelatin in PBS for 24 h at 4  C in agitation. In this and following steps, the incubations are protected from light exposure by wrapping the culture plate with aluminum foil. We use the following combinations of Alexa Fluor conjugated secondary antibodies, selected for the corresponding combinations of primary antibodies: – A: 488 anti-mouse, 555 anti-rabbit and 647 anti-guinea pig. – B: 405 anti-rabbit, 488 anti-mouse IgG2a, 555 anti-mouse IgG1, and 647 anti-guinea pig.

Vestibular Toxicity Assessment by Immunofluorescence l l

l

43

Rinse with PBS (2  10 min). For combination A only: Incubate with 1μL/mL DAPI in PBS (15 min). Rinse with PBS (4  5 min).

At this step, the vestibular epithelia can be mounted on slides as whole mounts or embedded for sectioning as explained in the next subsection. For mounting, we use Mowiol, but commercial mounting media suitable for immunofluorescence can also be used. 3.2 Post-Labeling Inclusion and Sectioning

– Embed epithelia in gelatin/albumin solution overnight at 4  C. – In an Eppendorf tube, start polymerization of a small volume (e.g., 250μL) of the gelatin/albumin solution by adding 9% (V/V) of 25% glutaraldehyde. Immediately transfer the mixture into a mold to form the bottom half of the block. Avoid the formation of bubbles. The solidification process can be slowed down if all the reagents are placed previously on ice. Cooling is recommended to facilitate the transfer of the polymerizing solution while avoiding the formation of bubbles. – The sample should be placed appropriately oriented on the gelatin/albumin solution (Fig. 3a) once the top layer of the solution is firm but still sticky (usually within 5–10 min after adding the glutaraldehyde to the gelatin/albumin solution). Noting the orientation of the epithelia is important because it will determine the orientation for sectioning (see Note 6). – Carefully absorb the excess liquid gelatin/albumin surrounding the sample with a paper towel. Pay special attention that movement of the gelatin/albumin solution does not shift the orientation of the epithelium and that the epithelium does not adhere to the paper towel. – Draw a diagram with the location and orientation of the sample in the block to facilitate orientation for later sectioning. – Cover the sample with a second polymerizing solution (Fig. 3b), prepared as described above, to form the upper half of the block. Pour the solution slowly down the walls of the mold to avoid undesired displacement of the sample. – Let this second layer completely solidify for about 30 min. – Take the solid block out of the mold and cut it into a rectangular pyramid that contains the sample oriented in the center and near the top (Fig. 3c) to allow transverse sections of the epithelia with the vibrating microtome. – Store the blocks overnight at 4  C in a 4% paraformaldehyde solution. Cooling and additional fixation increases their stiffness for better sectioning. If stored for 4 or more days, they risk becoming too rigid for proper sectioning.

44

Alberto F. Maroto et al.

A

C

B

D

E

Fig. 3 Schematic of the procedure for post-labeling embedding and sectioning of the vestibular epithelia. (a) A crista placed on top of the bottom half of the gelatin/albumin block. (b) A second half of the block is formed on top of the first half, so the crista gets included into the block at the limit between the two halves. (c) After complete solidification, the block is cut into a pyramid with the sample in the appropriate orientation. (d) The block is sectioned horizontally using the vibratome. (e) Correctly oriented specimens yield transverse sections of the epithelia that contain both central and peripheral regions of the end organ, crista, or macula

– Section the specimens at 40μm in a vibrating microtome (Fig. 3d, e). Mount the sections with the appropriate medium (e.g., Mowiol), coverslip, and store protected from light at 4  C until observation. 3.3 HC and Synapse Counts by Confocal Microscopy and Image Analysis

Sections are observed using a confocal microscope. For quantitative analysis, images are obtained from optimally oriented sections with the 63 (NA: 1.4) objective. For comparisons among groups of animals, the image acquisition settings must be maintained when observing the samples from the different groups of animals within the same batch. To avoid processing bias, the same number of samples from each experimental group must be processed in parallel. Generally, the entire epithelia can be scanned with a continuous Z-stack spanning around 25μm with an optical section thickness of around 0.3μm. Image processing software packages are used for the 3D visualization of stacks. To obtain separate counts of HCI and HCII, we use combination “A” of primary antibodies, secondary antibodies, and DAPI (Fig. 4). Using this combination, all cell nuclei in the epithelia are labeled with DAPI and the cytoplasm of all HCs is labeled with Myo7a. In addition, the Caspr1 label of the calyceal junctions identifies HCI, and the cytoplasmic calretinin label (overlapping with Myo7a) identifies HCII. The calretinin label of calyx-only afferents in the central areas of the receptors is clearly distinguished

Vestibular Toxicity Assessment by Immunofluorescence

A

B

C

D

HCII : Myo7aCalrenin

d

E

All HCs : Myo7a

HCI : Myo7aCaspr1

HCI Calyx only : Myo7a-Caspr1Calrenin

45

b

c

e

Fig. 4 Immunolabeling for HC counts in whole-mount preparations of the saccular macula. Images show labeling obtained with antibodies against Myo7a (red), Caspr1 (green), and calretinin (while), and the nuclear stain DAPI (blue). (A) General view of the striolar area. Other images are paired with letters to show a z-stack compression (capital letter) and a 3D reconstruction in higher magnification (lower case) of all HCs (B, b), HCI (C, c), HCII (D, d), and calyx-only complex afferents encasing two or three HCIs in the striolar zone (E, e). Scale bars ¼ 30μm in A; 10μm in B, C, D, E; 5μm in b, c, d, e

from that of HCIIs. Since the whole vestibular epithelia is to be analyzed, thick stacks are obtained. To manage these, it is useful to use the Imaris blend option to reduce the opacity of channels and ease the task of counting. This blend option can be found in the 3D visualization mode of Imaris. The filtered images can be counted later with the cell counter plugin of ImageJ. The “A” combination of primary antibodies is also useful to evaluate the integrity of the calyceal junction between HCIs and calyx afferents. The calyceal junction has been shown to be dismantled and rebuilt in rodent models of chronic ototoxicity and washout, consistent with the loss and recovery of function observed in the animals [6, 9]. Enduring damage of the calyces has also been recorded in models of partial ototoxic lesions [7]. For synaptic analysis, we use combination “B” of primary and secondary antibodies (Fig. 5). The Myo7a and calretinin labels allow identification of all HCs, HCII, and calyx-only afferent

46

Alberto F. Maroto et al.

A

F

G

Merge

B

C

Ribeye

PSD95

D

E

Calrenin

Myo7a

H

Fig. 5 Detail of a vestibular epithelium (crista) labeled for synapse counts. Confocal microscopy images are taken from 40μm thick transverse sections obtained using a vibrating microtome from a specimen embedded in a gelatin/albumin block. Four color images (a) are obtained with fluorescence channels (shown separate in b–e) adequate to the conjugate secondary antibodies recognizing the primary antibodies against ribeye (b; shown in red in a), PSD95 (c; shown in green in a), calretinin (d; shown in blue in a), and Myo7a (e; shown in white in a). Higher magnification of a raw 4-color image where pre-synaptic ribeye puncta (red, arrows) and post-synaptic PSD95 puncta (green, arrowhead) can be observed. (g) HCII with synaptic puncta as obtained by HCII segmentation using Imaris software. (h) Final 3D image obtained with Imaris of the same cells in G after cell and synaptic puncta recognition. Scale bars ¼ 25μm in a–e; 10μm f–h

terminals. Pre-synaptic and post-synaptic puncta are labeled with ribeye and PSD-95, respectively. The anti-ribeye antibody also recognizes the transcription factor CtBP2, encoded by the same gene, and additionally labels the HC nuclei. Synaptic analysis is more complex than HC counting, and we use a semi-automated approach in Imaris to facilitate counting and to reduce potential biases. First, the diameters of pre- and post-synaptic components (puncta) are obtained with the measurement tool in the 2D slice visualization. These values are introduced as parameters into the program to aid detection of the puncta. Next, the spot function of

Vestibular Toxicity Assessment by Immunofluorescence

47

the program is used in the 3D visualization mode. This function identifies the synaptic puncta based on the expected diameters measured previously for each fluorescence channel. Detection of puncta can be further optimized by adjusting the “quality threshold” setting of the spot function. Once selected, the acquisition parameters are maintained for the rest of the experiment to avoid bias during the analysis. To distinguish synapses on HCI and HCII, segmentation of the stack by each cell type is done before the spot analysis step. To this end, the cell functionality of Imaris is used. Adequate recognition of the cells is based on fluorochrome channels and cell diameters.

4

Notes 1. Fixative solutions, including paraformaldehyde and glutaraldehyde, are highly toxic and must be handled with protective equipment under a fume hood. 2. Colors shown in figures have been selected for best visualization and do not correspond to the emission colors of the fluorochromes. 3. Method to prepare the gelatin/albumin solution. In an Erlenmeyer flask, bring the PBS to a boil and dissolve the alimentary gelatin using a magnetic stir bar. Once dissolved, allow it to cool down to room temperature. Then, add the BSA in small fractions while stirring. Wait for each fraction to dissolve completely before adding the next fraction. Adding the BSA to a warm solution will cause it to denaturate, spoiling the mixture. After dissolving the albumin, add and dissolve the sucrose. The gelatin can be stored in aliquots at 20  C. After thawing, it can be stored at 4  C for up to 3 days. 4. The use of 4% Triton-X-100 for permeabilization of the vestibular epithelia exceeds largely the concentrations used for other tissues, but it has been demonstrated to be optimal for many antibodies in this tissue [16]. 5. Fish gelatin has been found to be more effective in reducing background noise than other blocking agents, such as donkey serum, in these immunohistochemical protocols [16]. 6. When making the gelatin/albumin blocks, the epithelia are placed flat and the pyramids made with the appropriate orientation to obtain transverse sections. Appropriate orientation of maculae will allow peri-striolar and peripheral regions (see Fig. 2) of the utricle or saccule to be obtained in the same sections. The orientation of the crista is critical, as only one orientation will permit representative transversal sections as shown in Fig. 5.

48

5

Alberto F. Maroto et al.

Discussion The two protocols detailed here yield 4-color images suitable for HC and synapse quantification using image processing programs. The first combination of antibodies and protocol has been established as optimal in our hands to quantify HCs, providing a better assessment than initial approaches such as the use of anti-calbindin and anti-calmodulin antibodies [34]. Nevertheless, other alternative antibodies can be used for this aim. Thus, the rabbit antitenascin antibody listed in Table 1 offers a good alternative to the anti-Caspr1 antibody to identify calyx endings and hence, HCI. Also, the goat anti-SPP1/ostepontin (Table 1) offers a good label of HCI. However, using the indicated rabbit polyclonal antitenascin requires colabeling with a mouse monoclonal antiMyo7a (Table 1). Labeling with the goat polyclonal anti-SPP1/ ostepontin is not compatible with the use of secondaries raised in goat. Although alternate hosts for these primary and secondary antibodies are available, the combination we suggest has been found to yield the best results. For a distinct identification of the central zone of the receptors, labeling with an anti-oncomodulin antibody (Table 1, Fig. 2) is a good approach [35]. Labeling of the calyx afferents with anti-Caspr-1 or antitenascin antibodies is useful to identify HCI cells, but the integrity of the calyces is also an endpoint of interest in ototoxicity studies [6, 7, 9]. For further characterization of the calyces, several choices are available, including anti-beta III tubulin (Tuj-1 clone) [7, 13], anti-KCNQ5 [6, 16], and anti-Na+/K+ ATPase alpha-3 subunit [36] antibodies. To quantify synaptic puncta, we previously used a rabbit anticalretinin antibody (Table 1), anti-ribeye and either anti-PSD95 or anti-GluA2 (Table 1) [6, 9]. This triple labeling protocol missed a positive label for HCIs, which were suboptimally delineated using background staining. The use of the guinea-pig anti-calretinin allows the simultaneous use of the rabbit anti-Myo7a to label all HCs. Although theoretically this protocol should still allow the use of the anti-PSD95 and the anti-GluA2 antibodies, the quadruple labeling protocol yielded poor GluA2 staining. Labeling the GluA2 subunit of the AMPA receptors is a more direct measure of the capacity for glutamatergic synaptic signaling than labeling the scaffolding protein PSD-95 of the post-synaptic density. However, loss of PSD-95 immunopuncta has been demonstrated to be a result of ototoxic damage, and therefore this endpoint is adequate for ototoxicity assessment.

Vestibular Toxicity Assessment by Immunofluorescence

49

Acknowledgments This study was supported by grants RTI2018-096452-B-I00 (Ministerio de Ciencia, Innovacio´n y Universidades, Agencia Estatal de Investigacio´n, Fondo Europeo de Desarrollo Regional, MCIU/AEI/FEDER, UE), and 2017 SGR 621 (Age`ncia de Gestio´ d’Ajuts Universitaris i de Recerca, Generalitat de Catalunya). E.A.G. was supported by the Secretaria d’Universitats i Recerca del Departament d’Economia i Coneixement de la Generalitat de Catalunya (FI-DGR 2014 Program) and by the Ministerio de Educa˜ a (FPU 2014). A.B.G is a Serracio´n, Cultura y Deporte de Espan Hu´nter fellow. The confocal microscopy studies were performed at the Scientific and Technological Centers of the University of Barcelona (CCiT-UB). We thank Benjamı´n Torrejon-Escribano for advice on confocal imaging. We also thank Lara Sedo´-Cabezo´n for her contributions to the establishment of these protocols within her PhD Thesis. References 1. Bronstein AM (2013) Oxford textbook of vertigo and imbalance. Oxford University Press, Oxford 2. Boyes WK, Pouyatos B, Llorens J (2020) Sensory function. In: Pope CN, Liu J (eds) An introduction to interdisciplinary toxicology. Academic Press—Elsevier, London, pp 245–260 3. Michalski N, Petit C (2015) Genetics of auditory mechano-electrical transduction. Pflugers Archiv 467:49–72 4. Schacht J, Talaska AE, Rybak LP (2012) Cisplatin and aminoglycoside antibiotics: hearing loss and its prevention. Anat Rec 295:1837–1850 5. Wilkerson BA, Artoni F, Lea C, Ritchie K, Ray CA, Bermingham-McDonogh O (2018) Effects of 3,30 -iminodipropionitrile on hair cell numbers in cristae of CBA/CaJ and C57BL/6J mice. J Assoc Res Otolaryngol 19:483–491 6. Sedo´-Cabezo´n L, Jedynak P, Boadas-Vaello P, Llorens J (2015) Transient alteration of the vestibular calyceal junction and synapse in response to chronic ototoxic insult in rats. Dis Model Mech 8:1323–1337 7. Sultemeier DR, Hoffman LF (2017) Partial aminoglycoside lesions in vestibular epithelia reveal broad sensory dysfunction associated with modest hair cell loss and afferent calyx retraction. Front Cell Neurosci 11:331 8. Cassel R, Bordiga P, Carcaud J, Simon F, Beraneck M, Le Gall A, Benoit A, Bouet V,

Philoxene B, Besnard S, Watabe I, Pericat D, Hautefort C, Assie A, Tonetto A, DyhrfjeldJohnsen J, Llorens J, Tighilet B, Chabbert C (2019) Morphological and functional correlates of vestibular synaptic deafferentation and repair in a mouse model of acute-onset vertigo. Dis Model Mech 12(7):pii: dmm039115 9. Greguske EA, Carreres-Pons M, Cutillas B, Boadas-Vaello P, Llorens J (2019) Calyx junction dismantlement and synaptic uncoupling precede hair cell extrusion in the vestibular sensory epithelium during sub-chronic 3,30 -iminodipropionitrile ototoxicity in the mouse. Arch Toxicol 93:417–434 10. Seoane A, Demeˆmes D, Llorens J (2001) Pathology of the rat vestibular sensory epithelia during subchronic 3,30 -iminodipropionitrile exposure: hair cells may not be the primary target of toxicity. Acta Neuropathol 102:339–348 11. Sousa AD, Andrade LR, Salles FT, Pillai AM, Buttermore ED, Bhat MA, Kachar B (2009) The septate junction protein caspr is required for structural support and retention of KCNQ4 at calyceal synapses of vestibular hair cells. J Neurosci 29:3103–3108 12. Moser T, Grabner CP, Schmitz F (2020) Sensory processing at ribbon synapses in the retina and the cochlea. Physiol Rev 100:103–144 13. Sadeghi SG, Pyott SJ, Yu Z, Glowatzki E (2014) Glutamatergic signaling at the vestibular hair cell calyx synapse. J Neurosci 34:14,536–14,550

50

Alberto F. Maroto et al.

14. Eatock RA, Songer JE (2011) Vestibular hair cells and afferents: two channels for head motion signals. Annu Rev Neurosci 34:501–534 15. Dechesne CJ, Winsky L, Kim HN, Goping G, Vu TD, Wenthold RJ, Jacobowitz DM (1991) Identification and ultrastructural localization of a calretinin-like calcium-binding protein (protein 10) in the Guinea pig and rat inner ear. Brain Res 560:139–148 16. Lysakowski A, Gaboyard-Niay S, CalinJageman I, Chatlani S, Price SD, Eatock RA (2011) Molecular microdomains in a sensory terminal, the vestibular calyx ending. J Neurosci 31:10,101–10,114 17. Burns JC, Kelly MC, Hoa M, Morell RJ, Kelley MW (2015) Single-cell RNA-Seq resolves cellular complexity in sensory organs from the neonatal inner ear. Nat Commun 6:8557 18. Aran JM, Erre JP, Guilhaume A, Aurousseau C (1982) The comparative ototoxicities of gentamicin, tobramycin and dibekacin in the Guinea pig. A functional and morphological cochlear and vestibular study. Acta Otolaryngol Suppl 390:1–30 19. Llorens J, Demeˆmes D, Sans A (1993) The behavioral syndrome caused by 3,30 -iminodipropionitrile and related nitriles in the rat is associated with degeneration of the vestibular sensory hair cells. Toxicol Appl Pharmacol 123:199–210 20. Lopez I, Honrubia V, Lee SC, Schoeman G, Beykirch K (1997) Quantification of the process of hair cell loss and recovery in the chinchilla crista ampullaris after gentamicin treatment. Int J Dev Neurosci 15:447–461 21. Maroto AF, Barrallo-Gimeno A, Llorens J. BioRxiv preprint. https://doi.org/10.1101/ 2020.12.21.423804 ˜ a-Ruı´z S, Boadas-Vaello P, Sed22. Saldan o´-Cabezo´n L, Llorens J (2013) Reduced systemic toxicity and preserved vestibular toxicity following co-treatment with nitriles and CYP2E1 inhibitors: a mouse model for hair cell loss. J Assoc Res Otolaryngol 14:661–671 23. Hasson T, Gillespie PG, Garcia JA, MacDonald RB, Zhao Y, Yee AG, Mooseker MS, Corey DP (1997) Unconventional myosins in inner-ear sensory epithelia. J Cell Biol 137:1287–1307 24. McInturff S, Burns JC, Kelley MW (2018) Characterization of spatial and temporal development of type I and type II hair cells in the mouse utricle using new cell-type-specific markers. Biol Open 7(11):pii: bio038083 25. Desmadryl G, Dechesne CJ (1992) Calretinin immunoreactivity in chinchilla and Guinea pig vestibular end organs characterizes the calyx

unit subpopulation. Exp Brain Res 89:105–108 26. Schmitz F, Ko¨nigstorfer A, Su¨dhof TC (2000) RIBEYE, a component of synaptic ribbons: a protein’s journey through evolution provides insight into synaptic ribbon function. Neuron 28:857–872 27. Chen X, Levy JM, Hou A, Winters C, Azzam R, Sousa AA, Leapman RD, Nicoll RA, Reese TS (2015) PSD-95 family MAGUKs are essential for anchoring AMPA and NMDA receptor complexes at the postsynaptic density. Proc Natl Acad Sci U S A 112:E6983–E6992 28. Braude JP, Vijayakumar S, Baumgarner K, Laurine R, Jones TA, Jones SM, Pyott SJ (2015) Deletion of Shank1 has minimal effects on the molecular composition and function of glutamatergic afferent postsynapses in the mouse inner ear. Hear Res 321:52–64 29. Sultemeier DR, Choy KR, Schweizer FE, Hoffman LF (2017) Spaceflight-induced synaptic modifications within hair cells of the mammalian utricle. J Neurophysiol 117:2163–2178 30. Niedzielski AS, Wenthold RJ (1995) Expression of AMPA, kainate, and NMDA receptor subunits in cochlear and vestibular ganglia. J Neurosci 15:2338–2353 31. Seoane A, Demeˆmes D, Llorens J (2003) Distal effects in a model of proximal axonopathy: 3,30 -iminodipropionitrile causes specific loss of neurofilaments in rat vestibular afferent endings. Acta Neuropathol 106:458–470 32. Gaboyard-Niay S, Travo C, Saleur A, Broussy A, Brugeaud A, Chabbert C (2016) Correlation between afferent rearrangements and behavioral deficits after local excitotoxic insult in the mammalian vestibule: a rat model of vertigo symptoms. Dis Model Mech 9:1181–1192 33. Gove C, Walmsley M, Nijjar S, Bertwistle D, Guille M, Partington G, Bomford A, Patient R (1997) Over-expression of GATA-6 in Xenopus embryos blocks differentiation of heart precursors. EMBO J 16:355–368 34. Cunningham LL (2006) The adult mouse utricle as an in vitro preparation for studies of ototoxic-drug-induced sensory hair cell death. Brain Res 1091:277–281 35. Hoffman LF, Choy KR, Sultemeier DR, Simmons DD (2018) Oncomodulin expression reveals new insights into the cellular organization of the murine utricle striola. J Assoc Res Otolaryngol 19:33–51 36. Schuth O, McLean WJ, Eatock RA, Pyott SJ (2014) Distribution of Na,K-ATPase α subunits in rat vestibular sensory epithelia. J Assoc Res Otolaryngol 15:739–754

Chapter 4 Morphometric Analysis of Axons and Dendrites as a Tool for Assessing Neurotoxicity Rhianna K. Morgan, Martin Schmuck, Ana Cristina Grodzki, Donald A. Bruun, Lauren E. Matelski, and Pamela J. Lein Abstract Chemical perturbation of the temporal or spatial aspects of axonal or dendritic growth is associated with neurobehavioral deficits in animal models, and structural changes in axons and dendrites are thought to contribute to clinical symptoms associated with diverse neurologic diseases. Consequently, axonal and dendritic morphology are often quantified as functionally relevant endpoints of neurotoxicity. Here, we discuss methods for visualizing and quantifying axonal and dendritic morphology of neurons from the peripheral or central nervous systems in in vitro and ex vivo preparations. These methods include visualization of neuronal cytoarchitecture by immunostaining axon- or dendrite-selective antigens, transfecting cells with cDNA encoding fluorescent proteins, or labeling cells using membrane permeable small molecules that distribute throughout the cytoplasm, Golgi staining or Diolistics, as well as quantifying axonal and dendritic morphology using semi-automated or fully automated image analysis. Key words Automated image analysis, Diolistics, Golgi staining, High-content imaging, Immunocytochemistry, LUHMES cells, Neurite outgrowth, Neurotoxicity, Primary neurons, Sholl analysis

1

Introduction Most neurons in the vertebrate central and peripheral nervous systems extend two types of processes—axons and dendrites— which differ functionally, biochemically, and structurally [1, 2]. The primary function of axons is to convey electrical signals from the neuronal cell body to downstream cells in the neural circuit. Axons typically extend considerable distances from the neuronal cell body to the target tissue, are of uniform caliber throughout most of their length, and are mostly unbranched until they reach the target tissue. In contrast, dendrites comprise the primary site of synaptic input to the neuron. Dendrites are broad at their proximal end and taper over their length to fine distal tips.

Rhianna K. Morgan and Martin Schmuck contributed equally to this work. Jordi Llorens and Marta Barenys (eds.), Experimental Neurotoxicology Methods, Neuromethods, vol. 172, https://doi.org/10.1007/978-1-0716-1637-6_4, © Springer Science+Business Media, LLC, part of Springer Nature 2021

51

52

Rhianna K. Morgan et al.

Dendrites tend to end proximal to the neuronal cell body and are highly branched throughout their length. Axons and dendrites are collectively referred to as neurites although the term “neurite” is also used to refer to neuronal processes that are not fully differentiated into an axon or dendrite, as occurs during early stages of neuronal differentiation [2], and is often the case with neuronal cell lines [3]. Axonal and dendritic morphology are critical determinants of neuronal function [1]. The shape of these processes affects signal processing within the neuron, while their number, length, and branching patterns determine the pattern of synaptic connections, which in turn regulates the distribution of information within the nervous system. Experimental evidence indicates that even subtle perturbations of temporal or spatial aspects of axonal and dendritic growth can cause persistent changes in synaptic patterning in the developing brain [4–7]. Clinical data confirm that altered axonal and dendritic structure is strongly associated with not only neurodevelopmental disorders but also neurodegenerative diseases [8– 13]. Based on such observations, chemical-induced changes in axonal and dendritic morphology are considered functionally relevant endpoints of neurotoxicity [14–16]. Thus, there is considerable interest in using morphometric analysis of axons and dendrites to screen chemicals for neurotoxic potential and to elucidate mechanisms of neurotoxicity. Multiple approaches are used to measure axonal and dendritic morphology, but across all approaches, there are two key steps: (1) visualizing the axonal plexus and/or dendritic arbor of neurons and (2) quantifying axonal and dendritic morphology. Choosing which method to use depends upon the experimental model, e.g., in vitro or ex vivo, and on whether distinguishing between toxic effects on axons vs. dendrites is a desired outcome [17–19]. In this chapter, we present several methods for visualizing and analyzing axonal and dendritic growth of peripheral and central neurons in vitro and ex vivo.

2

Materials

2.1 Immunocytochemical Approaches for Visualizing Axons and Dendrites In Vitro

To identify axons in primary peripheral neurons, antibodies (Ab) specific for protein gene product 9.5 (PGP9.5, Thermo Fisher Scientific 38–1000/Invitrogen PA110011, RRID:AB_1088162) or the phosphorylated forms of heavy (NF-H) and medium (NF-M) neurofilament subunits (SMI-31; Sternberger Immunocytochemicals/MilliporeSigma NE1022, RRID:AB_2043448) are widely used. For primary central neurons, Ab specific for tau-1 (MilliporeSigma MAB3420, RRID:AB_94855) is used to label axons. To label dendrites in primary neurons from either the peripheral or central nervous system, Ab specific for microtubule-

Morphometric Analysis of Axons and Dendrites

53

associated-protein-2 (MAP2) (SMI-52; Sternberger Immunocytochemicals/Synaptic Systems 188 004, RRID:AB_2138181/Invitrogen PA1-10005, RRID:AB_1076848) or non-phosphorylated forms of NF-H and NF-M neurofilament subunits (SMI-32; Sternberger Immunocytochemicals/SigmaAldrich 559844, RRID: AB_2877718) are effective. Additional materials needed include a fixative, usually 4% paraformaldehyde (MilliporeSigma) in 0.1 M phosphate buffer, and a permeabilization buffer [0.1% Triton-X100 (MilliporeSigma) in phosphate-buffered saline (PBS)] for antibodies to gain access to intracellular antigens. Also needed are a blocking buffer (PBS supplemented with 1–10% bovine serum albumin and/or 1–20% serum of the same species as the host species of the secondary antibody) to decrease binding of antibodies to nonspecific binding sites, fluorophore-tagged secondary antibodies that cross-react with the primary antibodies, and mounting medium [ProLong Gold Antifade Mountant (Invitrogen) for slides]. 2.2 Transfection of Cultured Neurons with Fluorescent Probes to Visualize Axonal or Dendritic Growth

Primary peripheral and central neurons can be transfected with low efficiency (which is desirable to facilitate visualization of the axonal plexus or dendritic arbor of individual neurons, particularly in high density cultures) using Lipofectamine 2000 (Invitrogen) following the manufacturer’s protocol. However, due to cytotoxic effects of lipid transfection reagents on primary neurons, we reduce the incubation time for transfection to only 1–2 h with total replacement of the transfection solution with cell culture media at the end of the incubation period. Dendrites are labeled by transfecting cultures with plasmids encoding microtubule-associated-protein-2B fused to either enhanced green fluorescent protein (MAP2BeGFP) or red fusion protein (MAP2B-FusRed), which is under the control of the neuron-specific CAG promoter [20]. Expression of these MAP2B fusion proteins is restricted to the somatodendritic compartment in cultured hippocampal neurons and does not alter their intrinsic dendritic growth patterns [20]. To label all processes (dendrites and axons), we transfect neurons with pCAG-tomato fluorescent protein (TFP) constructs. TFP distributes throughout the cytosol and does not alter neuronal morphology. We obtained these plasmids, which are not commercially available, from Dr. Gary Wayman (Washington State University, Pullman, WA, USA).

2.3 Live Cell Staining to Visualize All Neurites in the Culture

Calcein-AM (C3100MP, Molecular Probes), a cell-permeant dye used to determine cell viability in eukaryotic cells, will effectively fill all neurites in live cells. In live cells, the non-fluorescent calcein-AM is converted to fluorescent calcein-AM following acetoxymethyl ester hydrolysis by intracellular esterases. The fluorescent calceinAM can be measured at excitation/emission (ex/em) wavelengths

54

Rhianna K. Morgan et al.

of 494 and 517 nm, respectively. It is often used in conjunction with Hoechst 33342 (Thermo Fisher Scientific) and propidium iodide (PI; MilliporeSigma). Hoechst 33342 is a cell-permeable DNA stain that is excited by ultraviolet light and emits blue fluorescence at 460–490 nm. It preferentially binds adenine-thymine (A-T) regions of DNA and effectively labels nuclear chromatin. PI labels the nuclear membrane of compromised or dead cells and is used to identify non-viable cells in culture. 2.4 Diolistics to Label Processes of Neurons in Peripheral Ganglia or CNS Slice Cultures Ex Vivo

We use the Bio-Rad Helios Gene Gun Low-Pressure system for Diolistics labeling. This all-inclusive system contains a Tubing Prep Station, tubing cutter, cartridge storage vials and extractor tool, tungsten M-25 microcarriers, and Tefzel tubing. Additional supplies needed include polyvinyl pyrrolidone (PVP; MilliporeSigma), methylene chloride (Thermo Fisher Scientific), 1,10 -dioctadecyl-3,3,30 ,30 -tetramethylindocarbocyanine perchlorate (DiI) or other similar fluorescent dyes (Molecular Probes), hanging drop slides (Thermo Fisher Scientific), Sylgard 184 (Electron Microscopy Sciences), nitrogen, helium, three-way stopcock, ethanol, razor blades, and 35-mm plastic tissue culture dishes.

2.5 Golgi Staining to Visualize Dendrites in Intact Brain Tissue Ex Vivo

We use the FD Rapid GolgiStain kit (FD Neurotechnologies Inc.) for Golgi staining. Materials needed in addition to reagents supplied with this kit include plastic forceps to transfer tissues, plastic scintillation vials for incubation of tissue in Golgi solutions, 60 cm Whatman #1 paper to filter Golgi solutions, as well as standard histological supplies such as microscope coverslips and slides, staining dishes, slide holders, ethanol, chromium potassium sulfate, gelatin, sucrose (MilliporeSigma), and PBS.

3

Methods

3.1 Visualizing Neurites

Approaches for visualizing neurites include: (1) immunostaining for antigens selectively expressed in axons versus dendrites; (2) expressing cDNA encoding axon or dendrite-selective proteins linked to fluorescent tags or cDNA encoding intracellular fluorescent proteins that distribute throughout the entire cell; and (3) labeling neurons with cytoplasmic dyes or lipophilic membrane dyes (Table 1). Immunostaining for antigens selectively expressed in axons versus dendrites is most often used to visualize neurites in primary cell culture and offers the advantage of distinguishing axons from dendrites. This method labels all axonal or dendritic processes in the culture. Therefore, in low cell density cultures or at very early times after plating in higher density cultures, this approach can be used to identify processes extended by individual cells.

Morphometric Analysis of Axons and Dendrites

55

Table 1 Biomarkers for visualizing neurite outgrowth in vitro and ex vivo within the peripheral and central nervous systems Axons

Dendrites

In vitro: ICC analyses of phosphorylated Peripheral neurofilament (phospho-NF-H and nervous phospho-NF-M) subunits or tau as system, early as DIV 1 (3–8 cells/mm2) [31] e.g., postganglionic sympathetic neurons

In vitro: ICC analyses of MAP2 or dephosphorylated forms of M and H neurofilament subunits in cultures 48 h post-induction of dendritic growth (25,000 cells/cm2) [35, 36]

Ex vivo: IHC analyses of phosphorylated Ex vivo: Ballistic delivery system of fluorescent dyes (Diolistics) using neurofilament subunits or tyrosine 1,10 -dioctadecylhydroxylase (TH) in target tissues [72] 3,3,30 ,30 -tetramethyl-indocarbocyanine perchlorate (DiI)-coated tungsten beads [46, 50] Central nervous In vitro: ICC analysis of Tau-1 (33,000 cells/cm2) performed at DIV 2 [19] system, e.g., hippocampal neurons Ex vivo: “Brainbow” transgenic mice or injection of adenoviral or lentiviral vectors expressing cDNA encoding fluorescent proteins [21–23]

In vitro: MAP2B ICC or pCAGMAP2B-TFP transfection (83,000 cells/cm2) [19, 20] Ex vivo: Golgi staining [44]

ICC Immunocytochemistry, IHC Immunohistochemistry

Expressing cDNA encoding axon or dendrite-selective proteins linked to fluorescent tags or cDNA encoding intracellular fluorescent proteins that distribute throughout the entire cell are most useful when working with cultures of high cell density or mature cultures with extensive neurite outgrowth. In either case, it is often not possible to distinguish the dendritic arbor or axonal plexus of an individual cell from that of adjacent cells in the culture. The latter challenge can be overcome by using low efficiency transfection to label a small subpopulation of cells in the culture. Similarly, adenoviral or lentiviral infection with cDNA encoding fluorescent proteins can be used to label a subpopulation of neurons in the intact brain although this approach often involves infecting the living animal prior to harvesting of tissue for analysis [21–23]. Labeling neurons with cytoplasmic dyes or lipophilic membrane dyes can be used in vitro or ex vivo. Sometimes, there is incomplete labeling with this approach such that distal ends of processes are not labeled. Most membrane permeant cytoplasmic dyes and lipophilic membrane dyes label all neurites, so distinguishing axons from dendrites is only possible using structural criteria [2], which can be subjective. As with immunocytochemistry, in

56

Rhianna K. Morgan et al.

high density or mature low density cultures, it is challenging to identify the neurites associated with individual neurons. This issue can be overcome by measuring all neurites in a sample and then dividing by the number of nuclei (usually identified using a nuclear stain, such as Hoechst or DAPI). However, this approach is only feasible when cultures do not contain non-neuronal cells. Specific examples of these different approaches to visualize neurites are provided below. This is not an inclusive list, but rather a description of methods that have worked well in our hands. While we describe their application using specific models, these methods can be readily adapted to other models. 3.1.1 Visualizing Neurites in the LUHMES Neuronal Cell Line

Neuronal cell lines are significantly easier to obtain and maintain than primary neuronal cell cultures and are particularly useful when large numbers of cultures are needed (e.g., for screening chemical libraries) [24]. The laboratory of Marcel Leist (University of Konstanz, Germany) has championed the use of the human LUHMES (Lund Human Mesencephalic) cell line for neurotoxicity assays [25], and we have successfully used the LUHMES cell line to evaluate the effects of soluble immune mediators on neurite growth and retraction [26]. LUHMES cells were originally derived from 8-week-old female ventral mesencephalon. We obtained this cell line as a generous gift from the Leist laboratory, but LUHMES cells can also be purchased from the American Type Culture Collection (ATCC®; CRL-2927, RRID:CVCL_B056). This cell line is a sub-clone of the tetracycline-controlled, v-myc-overexpressing human mesencephalic-derived cell line MESC2.10 [27]. Like MESC2.10 cells, LUHMES cells can be differentiated into non-mitotic cells that recapitulate the morphological and biochemical characteristics of mature dopaminergic-like neurons by exposing undifferentiated, dividing LUHMES cells to tetracycline, glial cell line-derived neurotrophic factor (GDNF), and dibutyryl 3,50 -cyclic adenosine monophosphate (db-cAMP). LUHMES culture and differentiation are described in detail by Scholz and collaborators [3], and the quantification of neurite outgrowth using the Array-Scan II HCS Reader from Cellomics (Cellomics, PA, USA) has been described by Stiegler et al. [28]. Here, we describe measuring neurite outgrowth in LUHMES culture using a cell filling technique. While MAP2B immunocytochemistry has been used to image and quantify neurite growth in LUHMES cells [29], other studies have shown that the neurites extended by LUHMES cells express both dendritic and axonal properties [3]. Therefore, we generally use a cell filling technique to visualize neurites because it is quicker and less expensive. To assay neurite growth and retraction in LUHMES cells: l

Detach LUHMES cells from the substrate with trypsinization 48 h after differentiation begins (day 2).

Morphometric Analysis of Axons and Dendrites

57

l

Seed cells in 96-well plates at a density of 30,000 cells/well. Include both negative [cycloheximide (Sigma) at 3μM] and positive [p160ROCK inhibitor Y-27632 (Sigma) at 10μM] controls for neurite outgrowth in each 96-well plate [30]. These controls are most effective during the first 24 h after plating; when used later, they yield results that are more variable (Fig. 1).

l

Chemical exposures can begin one or more hours after seeding as processes start to generate within hours after re-plating, which is optimal for assessing the effects of chemicals on initial neurite outgrowth.

l

For neurite outgrowth exposures, prepare chemicals at twice the desired final concentration in tissue culture medium and add to each well at a volume equal to the volume of medium in the well (50μL) to minimize cell disruption.

l

Remove the medium from each well 24 h later and replace it with 1 Dulbecco’s PBS (with calcium and magnesium, ThermoFisher Scientific) containing 1μM calcein-AM, 1μg/mL Hoechst 33342 and 1.25μM PI.

l l

Incubate plates for 30 min at 37  C while protected from light. Image the cells immediately after incubation. In live cells, calcein-AM is converted to a green fluorescent cytoplasmic product that fills the soma and neurites and can be observed using filters with ex/em spectra of 494/517 nm. Hoechst 33342, a nuclear dye, is imaged using ex/em 350/461 nm filters. PI labels dead cells and is imaged using ex/em 535/617 nm filters.

This protocol can be modified to measure neurite retraction at later times during differentiation by exposing cells to chemicals on day 5 (3 days after plating) via a complete exchange of media and imaging on day 6. Brefeldin-A (Millipore Sigma), an inhibitor of membrane trafficking, causes significant neurite retraction and is used as a positive control (at 10μM) for neurite retraction assays. Chemical effects on neurite outgrowth can be influenced by plating surface, exposure time, and the method used for morphological analyses (Fig. 1). For example, 72, but not 48, h of exposure to PCB 95 (1 pM) increases neurite outgrowth in differentiated LUHMES cells plated on glass coverslips (see Note 1). Yet, neither a 72 nor a 48 h exposure to PCB 95 alters neurite outgrowth of LUHMES cells plated on tissue culture plastic. Moreover, the neurite growth promoting effects of PCB 95 were observed when cultures were immunostained for either MAP2B, a dendriteselective cytoskeletal protein, or for phosphorylated neurofilaments, an axon-selective antigen, but not when using calcein-AM and Hoechst 33342. Since MAP2B is expressed only in more mature LUHMES cells, while neurites are extended by both mature and immature LUHMES cells [3], these data suggest that the

58

Rhianna K. Morgan et al.

Fig. 1 Substrate, exposure time, and method of morphometric analysis influence outcome in cultured LUHMES cells, a human neuronal cell line. (a, b) LUHMES cells at similar passage number were plated at the same cell density onto (a) tissue culture plastic or (b) glass coverslips and exposed to vehicle (0.1% DMSO) or varying concentrations of PCB 95 for 24, 48, or 72 h. Y-27632, a p160ROCK inhibitor that increases neurite outgrowth during the first 24 h after plating, was used as a positive control. (Note: as indicated in this Figure, when used at later times after plating, Y-27632 does not increase neurite outrgrowth). Neurite outgrowth was imaged by labeling cells with calcein-AM. (c, d) LUHMES cells grown on glass coverslips were exposed to vehicle or PCB 95 for 24–72 h, then immunostained for axons or dendrites using antibodies specific for (c) phosphorylated neurofilaments or (d) MAP2B, respectively. Data are presented as the mean  SD (n ¼ 3 independent experiments with eight wells in each experiment). *Significantly different from DMSO vehicle control at p < 0.05; **p < 0.01 as determined by one-way ANOVA with Dunnett’s post hoc test; #significantly different from DMSO vehicle control as determined by Student’s t-test

Morphometric Analysis of Axons and Dendrites

59

morphogenic effects of PCB 95 are manifest only at latter stages of neuronal differentiation. The observation that immunostaining for either a dendrite or axon-selective antigen yields the same outcome is consistent with the observation that the neurites extended by LUHMES cells exhibit both dendritic and axonal properties [3]. 3.1.2 Labeling Axons Versus Dendrites in Dissociated Primary Sympathetic Neurons

Primary cultures of sympathetic neurons dissociated from the superior cervical ganglia (SCG) of laboratory rodents are a robust model system for assessing chemical effects on axonal or dendritic growth [17, 31]. SCG are easily accessible at varying life stages of rodents and yield a homogeneous neuronal cell population consisting of principal sympathetic neurons that can be maintained for weeks to months in serum-free defined culture medium in the absence or presence of ganglionic glial cells [31]. Primary sympathetic neurons extend a single process that is axonal within 24 h of plating, and this unipolar morphology can be sustained for up to 3 months when cultures are maintained in serum-free medium in the absence of glial cells [32, 33]. Robust dendritic growth can be induced in primary sympathetic neurons cultured in serum-free media by co-culturing them with ganglionic glial cells [34]. In sympathetic neurons grown in the absence of glial cells, dendritic growth can be triggered by the addition of recombinant BMP-2, 4, 6 or 7 (30–100 ng/mL; R&D Systems) or Matrigel (50–75μg/mL, Corning Life Sciences) to the culture medium [35]. The processes induced by BMPs and Matrigel appear about 48 h post-exposure, and exhibit functional, biochemical, and morphological characteristics of dendrites [33, 36]. The protocol for setting up primary sympathetic neurons from the SCG of perinatal rat or mouse pups on glass coverslips (ideal for morphometric analyses that can be done as early as 2 days in vitro (DIV)) has been published [35]. Protocols for immunostaining these cultures have also been published [37, 38] and are briefly described below: l

Fix cultures with 4% paraformaldehyde in 0.2 M phosphate buffer for 10 min at room temperature.

l

Permeabilize cultures with 0.1% Triton-X-100 in PBS for 10 min at room temperature.

l

Incubate cells in blocking buffer (PBS with 10% bovine serum albumin) for at least 1 h at room temperature.

l

To identify axons, incubate cultures for 1 h at room temperature or overnight at 4  C with Ab specific for phosphorylated NF-H and NF-M neurofilament subunits or Ab that recognizes synaptophysin. To visualize dendrites, incubate cultures with Ab specific for MAP2B or with Ab that recognizes dephosphorylated NF-H and NF-M. Dilute antibodies in blocking buffer (see Note 2).

60

Rhianna K. Morgan et al.

3.1.3 Labeling Axons and Dendrites in Dissociated Cultures of Primary CNS Neurons

l

Remove excess primary antibody by rinsing the cultures three times with PBS over 5 min.

l

Incubate with a fluorescent-tagged secondary antibody that recognizes the primary antibody for 1 h at room temperature or overnight at 4  C in the dark.

l

Remove excess secondary antibody by gently rinsing cultures three times with PBS over 5 min.

l

If grown on glass coverslips, cultures are mounted cell side down in mounting medium on glass slides and the edges sealed with nail polish.

l

If grown in multi-well plastic tissue culture plates, PBS containing DAPI is added to the well for 15 min, the wells rinsed twice with PBS, and the culture dish sealed with Parafilm™ to prevent evaporation until cultures are imaged.

Primary hippocampal and cortical neurons dissociated from perinatal rat or mouse pups are widely used in studies of neuronal morphogenesis and the effects of chemicals on axons and dendrites. In culture, these neurons exhibit a characteristic morphogenic program [39–41]. Stage 1 begins immediately after plating, when lamellipodia propagate around the somata signaling initial neurite formation. Growth cones develop in Stage 2 at the tips of neurites, enabling neurites to extend or retract short distances. Stage 2 lasts 12–36 h until one neurite begins to extend rapidly and its length exceeds that of the other “minor” neurites by 2- or 3-fold [40]. This single long neurite differentiates into the axon, while the other minor neurites slowly develop dendritic characteristics. As the culture ages, dendrites elongate and branch, and by approximately 21 DIV, synapses begin to form and dendritic spines appear [42]. This predictable morphogenic program makes primary hippocampal and cortical neurons a robust in vitro model for assessing chemical effects on axonal and dendritic morphologies of central neurons at varying stages of neuronal maturation. Primary cortical and hippocampal neurons can be dissociated from embryonic day 18 or postnatal day 0 or 1 mouse or rat neocortices or hippocampi, as previously described [19, 43]. The former will contain significantly fewer glial cells than the latter. The neurons can be maintained at very low density for long periods of time using the “Banker” method of inverting the coverslip on which the neurons are plated over a monolayer of astrocytes [43]. Alternatively, cortical and hippocampal neurons can be cultured at high density in the presence of endogenous glia (mostly astrocytes) on the same coverslip [19]. With either method, cortical and hippocampal neurons develop a morphology that resembles that of their in vivo counterparts [20, 43]; however, we have found that the morphogenic response to neurotoxic chemicals can vary

Morphometric Analysis of Axons and Dendrites

61

depending on which culture method is used. In our experience, the chemical-induced morphogenic response of neurons cultured as high density neuron-glia co-cultures more closely resembles the response observed in vivo [14]. Additionally, the substrate on which the cells are plated can influence the morphogenic response to chemicals. Plating on glass coverslips is the best practice because lipophilic compounds, such as PCB 95, stick to plastic, which prevents or minimizes their effect on neuronal morphogenesis (Fig. 2). To quantify axons in primary hippocampal or cortical cultures: l

Plate hippocampal or cortical neuron-glia co-cultures at 33,000 cells/cm2.

l

To quantify chemical effects on the complete axonal plexus of individual neurons, expose cells to vehicle or chemicals for 48 h beginning at 4 h post-plating [38].

l

Fix cultures with 4% paraformaldehyde in 0.2 M phosphate buffer for 10 min at room temperature.

l

Permeabilize cultures with 0.25% Triton-X-100 in PBS for 5 min at room temperature.

l

Incubate cells in blocking buffer (PBS with 10% bovine serum albumin and/or 10% goat serum) for at least 1 h at room temperature or overnight at 4  C.

l

To identify axons, incubate cultures for 1 h at room temperature with Ab specific for tau-1. Dilute antibodies in blocking buffer (see Note 2).

l

Remove excess primary antibody by rinsing the cultures three times with PBS over 5 min.

l

Incubate with a fluorescent-tagged secondary antibody that recognizes the primary antibody for 1 h at room temperature or overnight at 4  C in the dark.

l

Remove excess secondary antibody by gently rinsing cultures two times with PBS over 5 min.

l

If grown on glass coverslips, cultures are mounted cell side down in mounting medium on glass slides and the edges sealed with nail polish.

l

If grown in multi-well plastic tissue culture plates, PBS containing Hoechst or DAPI is added to the well for 15 min, wells are rinsed twice with PBS, and the culture dish sealed with Parafilm™ to prevent evaporation until cultures are imaged.

To quantify dendritic arborization (total length, number of terminal tips, neurite mass, and number of branching points) in high density neuron-glia co-cultures, transfect a subpopulation of neurons (0.1–0.5%) with cDNA encoding a fluorescent protein:

Fig. 2 Substrate composition influences morphogenic response of primary rat central neurons to neurotoxic agents. Primary neuron-glia cultures dissociated from perinatal rat hippocampi and cortices were established at high density and maintained as previously described [19] on either tissue culture plastic or glass coverslips. At day in vitro (DIV) 6, cultures were transfected with cDNA encoding a MAP2B-eGFP construct to label the dendritic arbors of approximately 10% of the neurons in the culture. At DIV 7, cultures were exposed to vehicle (0.1% DMSO) or PCB 95 at varying concentrations for 48 h. Dendritic arborization was quantified in eGFP-positive cells (10–30 cells per well) as the total dendritic length and the total number of dendritic tips per neuron. Data are presented as the mean  SD (n ¼ 5–9 wells from three independent dissections). *Significantly different from DMSO vehicle control at p < 0.05; **p < 0.01; ***p < 0.001 as determined by one-way ANOVA with Dunnett’s post hoc test

62 Rhianna K. Morgan et al.

Morphometric Analysis of Axons and Dendrites l

l

3.1.4 Ex Vivo Labeling of Dendritic Arbors in Peripheral Autonomic Ganglia Using Diolistics

63

Dissociate cells from the neocortices or hippocampi of postnatal mice or rats. These can be sex segregated [44]. Plate cells on glass coverslips at 83,000 cells/cm2.

l

At 4 DIV, treat cells with the anti-mitotic agent, cytosine arabinoside (Ara-C; Sigma) to curb glial cell proliferation by replacing 50% of the conditioned media with medium supplemented with 5μM Ara-C to yield a final Ara-C concentration of 2.5μM.

l

The optimal window to assess the effects of chemicals on dendritic growth in high density neuron-glia co-cultures is between 4–10 DIV because this is the period of peak dendritic growth [20] although chemicals can be added to more mature cultures to assess effects on dendritic retraction.

l

Transfect cells with plasmid encoding MAP2B-eGFP, MAP2BFusRed, or pCAG-TFP [20] using Lipofectamine 2000 on DIV 6. Exceptions to the manufacturer’s protocol include leaving the transfection solution on the cells for only 1–2 h and immediately replacing it at the end of the transfection period with conditioned media.

l

On 7 DIV, treat cultures with vehicle or varying concentrations of test chemicals for 48–72 h [45].

l

Fix cultures with 4% paraformaldehyde on 9 DIV and mount to glass slides using ProLong Gold Antifade reagent with DAPI (Thermo Fisher Scientific).

Diolistics uses pressure to deliver small beads coated with lipophilic carbocyanine dyes to stochastically label a subpopulation of cells within intact neural tissues [46]. The labeling of individual cells is rapid such that the entire dendritic arbor of neurons can be visualized within minutes after particles contact the cell membrane. This method is a variant of biolistics in which plasmids are delivered into hard-to-transfect cells, such as plant cells, by delivering small gold or tungsten particles coated with plasmids or other cDNA material using a gene gun. Diolistics was first described by Gan et al. [47] and uses fluorescent lipophilic dyes, such as DiI (1–10 -dioctade-cyl3,3.30 ,30 -tetramethylindocarbocyanine perchlorate, Thermo Fisher Scientific). Because of their lipophilic nature, these dyes partition into the plasma membrane, thereby outlining neuronal processes and spines. Membrane staining is more efficient and allows for better visualization of small thin protrusions than cytoplasmic staining. DiI-labeled neurons can be observed by high-resolution imaging, such as confocal or two-photon microscopy, and can be digitally reconstructed in precise detail [48, 49]. Our laboratory has used Diolistics to successfully label individual neurons in intact rat and mouse superior cervical ganglia (SCG) using the Helios® Gene Gun system from Bio-Rad [50]; we and

64

Rhianna K. Morgan et al.

others have used this technique to label individual neurons in CNS organotypic slice cultures and cultured cells [20, 51]. The “bullets” used with the Helios gene gun are made of sections of Tefzel tubing filled with tungsten beads pre-coated with DiI. To coat the Tefzel tubing with polyvinylpyrrolidone (PVP) (Sigma) to improve adherence of the tungsten beads to the interior walls of the tubing: l

Prepare a stock PVP solution at 20 mg/mL in 100% ethanol. Each day that bullets are prepared, the PVP solution is further diluted to 0.1 mg/mL in 100% ethanol.

l

Cut approximately 30 inches of Tefzel tubing (Bio-Rad), rinse with 100% ethanol, and feed it into the Bio-Rad Prep station stopping just before the tubing enters the prep station gas outlet.

l

Use a 10 mL syringe with a three-way stopcock mounted vertically on a ring stand next to the tubing prep station to avoid inserting air bubbles into the tubing and a short piece of silicone tubing to connect the stopcock to the Tefzel tubing.

l

Pour the PVP solution into the syringe and open the stopcock to allow the PVP solution to fill the Tefzel tubing.

l

Once the tubing is filled, close the stopcock to keep the PVP solution in the tubing for 5 min. Do not overfill the tubing and blot any excess with a Kimwipe™.

l

Withdraw the PVP solution using the syringe.

l

Gently push the Tefzel tubing into the gas outlet and connect it to the barbed hose connector on the prep station connected to a dry nitrogen gas source. Dry the tubing using a nitrogen gas rate of 0.4 L/min for approximately 5 min. To prepare the tungsten particles:

l

Suspend 15 mg of 1.1μm tungsten M-17 particles (Bio-Rad) in 150μL methylene chloride (99.9% pure, Fisher Scientific) in a 1.5 mL microfuge tube. Mix well and sonicate for 3 min to break up any bead clumps.

l

Pipette the suspended particles onto a glass hanging drop slide and air dry for approximately 10–15 s to allow for even distribution of the particles.

l

Place 3 mg of DiI in a microfuge tube, add 100μL of methylene chloride, and vortex.

l

Pipette solution on top of the dried tungsten particles. Dried particles ( f1), continuous and closely spaced in frequency, are presented to the ear canal, the acoustic response of the cochlea is called a Distortion-Product Otoacoustic Emission (DPOAE). When the two traveling waves induced by these two frequencies interact, they produce a family of intermodulation distortions at different mathematically related frequencies, e.g., the sum and difference of f1 and f2, as well as the sums and differences of multiples of those frequencies. In addition to the distortion phenomenon, a coherent-reflection mechanism also occurs. These complex mechanisms are not the topic of the current chapter and detailed descriptions of the physics of DPOAE generation can be found elsewhere [12]. The most prominent DPOAE that can be measured in both humans and laboratory rodents is often the cubic distortion product 2*f1
f2, which explains why it has been the most used in clinical settings and laboratory experiments. The chapter focuses on the description of a method to record cubic DPOAEs in rats. However, it is worth mentioning that an excellent methodological description of DPOAE recordings in mice is also available [13]. Actually, DPOAEs have been measured in almost all common and exotic laboratory species, including amphibians and birds. All DPOAE measurement systems are composed of four parts: 1. a frequency synthesizer, coupled to an amplifier, which produces the sinuses for the two pure tones. 2. two speakers, which transduce those frequencies. 3. one microphone, also coupled to an amplifier, which converts the cochlear response into an electrical signal. 4. a spectral analyzer, which allows the averaging and the Fourier transform of electrical signal. The critical step in recording DPOAEs is the separation of the meaningful DPOAE response from the acoustic noise coming from the environment and the animal. This separation requires the averaging of numerous acquisitions, a pertinent choice of f1 and f2 settings, as well as a proper setup of spectral analysis parameters. Since the reliability of DPOAE measurement depends mostly on these parameters, they are detailed in this chapter. Although DPOAEs can be, in theory, measured in non-anesthetized animals, using habituation and contention (e.g., guinea pigs) or a surgically implanted cranial anchor (e.g., rats), the procedure is much easier and the DPOAEs more stable in ketamine/xylazine anesthetized animals. In addition, the ketamine/

How to Measure Distortion Product Otoacoustic Emissions in Rats?

93

xylazine mixture does not significantly alter DPOAE levels, as opposed to isoflurane [14]. It is important to bear in mind that the OHC function of the low-frequency region of the cochlea (apex) cannot be investigated using DPOAEs. Indeed, the gain of the cochlear amplifier is only a few-fold at the apex of the cochlea against a 1000-fold at the base [15]. Consequently, DPOAEs can only be considered as reliable with f2 > 3000 Hz in rats. The rat being able to hear sounds up to 60 kHz, we do not advise to use human commercial DPOAE systems dedicated to the clinic, which are generally limited to 8 kHz. However, some off-the-shelf systems specifically designed for the measurement of DPOAE in rats are commercially available (see Note 1).

2

Materials l

Otoscope.

l

Ketamine.

l

Xylazine.

l

Data acquisition unit (e.g., Bru¨el & Kjær Type 3560D).

l

Two high-sampling, low-distortion signal generators (e.g., Bru¨el & Kjær Type 3110).

l

High-frequency emitters (e.g., Bru¨el & Kjær Type 4192 microphone capsules) to produce pure tones up to 60 kHz.

l

Low-noise microphone (e.g., Knowles FG-23329-CO5).

l

Low-distortion amplifier(s) (e.g., Femto DLPVA-100-B Series).

l

PC equipped with a software to control the data acquisition system (e.g., Bru¨el & Kjær PULSE interface).

l

Multifunction Type 4226).

l

Sound-attenuating chamber (see Note 2).

l

Heating pad (e.g., Harvard Apparatus).

l

1/8 inch reference microphone (e.g., Bru¨el & Kjær Type 4138).

l

Home-made cylindrical calibration cavity (0.045 cm3).

l

Small curved forceps and elastic bands.

l

Laboratory stand.

l

Articulated holder.

acoustic

calibrator

(e.g.,

Bru¨el

&

Kjær

94

3

Thomas Venet et al.

Methods

3.1 Equipment Setup (Fig. 1) 3.1.1 Generators and Analyzer

DPOAE measurements are performed using a Bru¨el & Kjær Type 3560D data acquisition unit equipped with two 3110 modules, which offer two signal generators with 204.8 kHz bandwidths and 24-bit D/A convertors, as well as two measurement channels (see Note 3). This configuration is therefore perfectly suited for measuring rat’s hearing whose audible range extends up to 60 kHz. PULSE software can be controlled by a Visual Basic OLE interface to design automatized measurement sequences. The use of this interface speeds up the measurement and increases the reliability of the data recording. Calibration and periodic checks carried out with a multifunction acoustic calibrator (Bru¨el & Kjær Type 4226) guarantee a maximum error of 0.15 dB up to 4000 Hz and 0.20 dB beyond.

3.1.2 DPOAE Measurement Probe

Most DPOAE measurement probes available on the market are dedicated to humans. The upper frequency limit is generally less than 8000 Hz, which is insufficient for the rat. Therefore, we manufactured a probe in the laboratory, which is detailed hereafter (Fig. 2). This probe uses two and a half inch microphone capsules (Bru¨el & Kjær Type 4192) as a sound transmitter for f1 and f2. These capsules require a 200 V polarization, which is superimposed on the sinusoidal signal in a conditioner near the probe. Two cascaded amplifiers (voltage gains: 20 + 5 dB) amplify the sinusoidal signal, allowing an alternative effective voltage of 45 Vrms to be reached. The usable bandwidth ranges from 500 to 30,000 Hz. The microphone used to measure the sound signal is a Knowles FG 23329C05 miniature microphone. The signal delivered by this microphone is amplified by 40 dB by a conditioner placed near the probe (homemade voltage amplifier whose characteristics are close to those of commercial amplifiers; see Subheading 2 above) and operating on battery (12 V). The bandwidth of this microphone ranges from 100 to 10,000 Hz (3 dB). However, by judiciously choosing the frequencies, this custom probe fitted with this microphone allows DPOAE measurements up to 25,000 Hz after calibration. Beyond that, the sensitivity of the assembly (probe + microphone) is too low and this device is no more usable. The different elements of this device (probe, conditioner, amplifiers) are periodically checked and calibrated. The quality of the electrical connections and even more of the mechanical

How to Measure Distortion Product Otoacoustic Emissions in Rats?

95

Fig. 1 Schematic description of the DPOAE measurement system for the rat

assembly have a great influence on the performance of the system, especially at high frequencies (see Note 4). 3.1.3 Probe Calibration and Periodic Verification

In order to calibrate the sound emissions for the rat’s hearing system, a cavity with a volume equivalent to that of the ear canal is used (Fig. 3). For the adult (6-month-old) Brown Norway and Long-Evans type rat, the volume is approximately 0.045 cm3. The custom-made cylindrical calibration cavity (length ¼ 6.5 mm; diameter ¼ 2.9 mm) allows the insertion of the probe to be fitted with an audiometric plug with a diameter of 6 mm at one end. The other side of the cavity is fitted with a Teflon ring, which allows the reference microphone to be inserted (Bru¨el & Kjær type 4138, 1/8 inch). The probe is calibrated between 500 and 30,000 Hz. The probe calibration allows you to define the levels emitted in dB

96

Thomas Venet et al.

Fig. 2 Probe diagram

Fig. 3 [Probe/cavity/reference microphone] assembly for calibration

SPL by the f1 and f2 transmitters (see Note 5). It allows choosing the frequencies that can be used according to the response curve measured at the reference microphone and at the microphone of the probe. A simplified procedure has been defined for a rapid verification that can be carried out just before each series of recordings. This verification consists in launching a DPgram measurement (see Subheading 3.3.2), leaving the probe free (without rat, cavity, and audiometric plug). For each f1 and f2 frequencies of the DPgram,

How to Measure Distortion Product Otoacoustic Emissions in Rats?

97

the levels L1 and L2 measured by the probe microphone are compared to the reference levels measured under the same conditions following calibration of the probe. The maximum permissible error is fixed at 2 dB. This cavity-free procedure provides good reproducibility even at high frequencies. 3.2 DPOAE Measurements 3.2.1 Animal Preparation and Experimental Setup

General anesthesia is induced by an intraperitoneal (i.p.) injection of a mixture of Ketamine (Clorketam®, 45 mg/kg) and Xylazine (Rompun®, 5 mg/kg). Otoscopic examination has to be performed before the measurements in order to ensure the absence of infection, obstruction of the external auditory canal, or anomaly of the tympanic membrane. It is very important to stabilize the rat’s body temperature at 38  C throughout the whole procedure using a heating blanket (see Note 6). Additionally, the maintenance of a steady atmospheric pressure around the animal is critical because anesthetized animals cannot regulate the middle ear pressure and the tympanic tension strongly influences the DPOAE level (see Note 7). On the bench of the soundproof chamber, place a laboratory stand with a horizontal rod and an articulated holder on which you can secure the DPOAE probe. This holder needs to be easily adjustable for the convenient placement of the probe into the ear canal of the animal. Be sure to have two small clamps linked to two elastics bands at your disposal. Place a 6-mm diameter audiometric plug on the probe.

3.2.2 Probe Placement

Place the anesthetized rat on its side (with the test ear facing up) on the heating blanket. Approach the probe until it touches the entrance of the ear canal. Use the small clamps to gently pull up the pinna while carefully inserting the probe microphone (Fig. 4). Note that it is better to pull up the pinna than to push down the probe.

3.2.3 Stimulation Parameters

Table 1 summarizes the parameters used for the primary frequencies f1, f2, and their respective intensities L1 and L2. The DPOAE application should allow two types of paradigms, DPgram or input-output (I/O). The DPgram is a plot of DPOAE amplitude as a function of stimulus frequency with the stimulus level held constant. By contrast, the I/O consists of DPOAE amplitude as a function of the level of the stimulus for a given frequency. It is also convenient to be able to perform more complex protocols using a mix of DPgrams and I/Os, or to monitor the time course of a DPOAE response using a single frequency/level combination. The optimal stimulation parameters yielding maximum DPOAEs have been defined experimentally by adjusting the intensity difference between L1 and L2 for Long-Evans (L1 ¼ L2 + 14 dB) and Brown Norway (L1 ¼ L2 + 11 dB) rats. They might differ

98

Thomas Venet et al.

Fig. 4 Brown Norway (a) and Long-Evans (b) rats during DPOAE measurements Table 1 Value of the main parameters used for DPOAE measurements in rats Frequency ratio ( f2/f1)

1.2

Level difference (L1
L2)

11–14 dB depending on the strain

Frequency range ( f2)

3600–25,000 Hz

Intensity range (L1)

25–70 dB

slightly in your specific conditions. Indeed, a value of 10 dB (L1 ¼ L2 + 10 dB) is often mentioned in the literature. Alternatively, a symmetric protocol can be applied, in which the primary tones, f1 and f2, are presented at equal levels (L1 ¼ L2). The best practice is to try out several intensity parameters and to select the one that yields the optimal DPOAE amplitudes. Our standard measurement protocol includes five frequencies: f2 ¼ 3600 Hz, f2 ¼ 6000 Hz, f2 ¼ 9600 Hz, f2 ¼ 17,520 Hz, f2 ¼ 25,440 Hz. The ratio f1/f2 is fixed at 1.2. Because the origin of DPOAEs on the organ of Corti is mainly located near the f2 frequency and because DPOAE values depend primarily on the L1 level, our DPOAE measurements are expressed and plotted as a function of the f2 frequency and L1 intensity. 3.2.4 DPOAE Signal Acquisition

The acoustic signal to be analyzed consists of three periodic signals: l

l

f1 and f2 are the sinusoidal signals emitted by the probe to generate DPOAEs in the inner ear. fDPOAE ¼ 2*f1
f2 is the sinusoidal signal corresponding to the cubic product of acoustic distortion emitted by the inner ear and transduced by the middle ear (see Note 8).

How to Measure Distortion Product Otoacoustic Emissions in Rats?

99

These three signals are superimposed over a background noise. This background noise (or random noise) is produced by the ambient acoustics of the room and the endogenous noises of the animal. It mainly consists of low and medium frequencies. At high frequencies, electronic background noise is preponderant. Background noise can also contain periodic signals such as electromagnetic interference (telephone, chopper circuits on power lines). The measured signal is analyzed by fast Fourier transformation algorithms (real-time FFT) on Bru¨el & Kjær analyzers. The acquisition parameters, useful frequency bandwidth, and the number of lines of the spectrum are linked by the following formula: N ¼ 2:56 N line F e ¼ 2:56 F span dF ¼ F span =N line T ¼ NF e with Nline

Number of FFT lines composing the frequency spectrum

Fspan

Maximum frequency of the frequency analysis bandwidth (Hz)

N

Number of acquisition block samples

Fe

Sampling frequency (Hz)

dF

Frequency resolution (Hz)

T

Duration of an acquisition block (s)

The type of average (temporal or spectral, linear, or exponential), the number of time blocks averaged as well as the temporal and frequency weights are adapted according to the type of measurement carried out: DPgram, I/O or follow-up of the time course. The acquisition parameters common to the different types of measurement are as follows: l

l

Average acquisition blocks The averages are calculated on the auto-spectrum. This corresponds to the average of the signal energy. This type of analysis provides a good representation of the total energy contained in the signal, including random signals composing the background noise, which is important for judging the signal/noise ratio of the measurement. A temporal average of the blocks would not allow a good evaluation of the background noise since only the signals in phase with the acquisition frequency of the blocks are correctly taken into account. Time weighting of acquisition blocks

100

Thomas Venet et al.

Fig. 5 FFT analysis with sampling frequency Fe ¼ 65,530 Hz. Size and period of the time block: 8192 points/ 125 ms. Number of FFT lines: 3200. Frequency resolution: 8 Hz. (a) Effect of Hanning (dashed line) vs. rectangle (dotted line) windowing on a sinusoidal signal of 1 Vrms at 1004 Hz. (b) Picket fence effect on a sinusoidal signal of 1 Vrms at 1000 or 1004 Hz. Solid curve, signal at 1000 Hz or 125 periods per time block. Dashed curve, signal at 1004 Hz or 125.5 periods per time block

3.2.5 Specific Signal Processing Parameters

Hann windowing (Fig. 5a) is systematically used with an overlap of 66% between blocks. This analysis is well suited for continuous and pseudo-random signals. The Hann window eliminates the edge effects of the acquisition blocks, which gives good frequency resolution. The 66% overlap is optimal for the Hann window because it allows for uniform overall weighting without losing time data (Fig. 6). Without overlap, in case an event occurred at the junction of two blocks, its intensity would be attenuated and the contained information lost. l Choice of frequencies By picket fence effect, an FFT analysis can also cause an error related to digitization depending on the position of the true frequency of the phenomenon analyzed in relation to the frequency resolution of the FFT. With a Hann window, this error can be up to 1.4 dB (Fig. 5b). To avoid this problem, the DPOAE application adjusts the frequencies f1 and f2 so that the fDPOAE is a multiple of the frequency resolution of the FFT. Depending on the type of measurement performed, the bandwidth (sampling frequency), the number of lines (frequency resolution) and the number of averages are different. Table 2 summarizes the configurations suitable for obtaining an average value (DPgram or I/O) or monitoring the DPOAE responses over time. The rejection of the FFT spectra disturbed by the background noise is carried out in real time. The DPOAE application rejects any acquisition according to the following criterion:

How to Measure Distortion Product Otoacoustic Emissions in Rats?

101

Fig. 6 Effect of overlap on the weight of time samples with Hann windowing

Table 2 Parameters of the FFT analysis for DPgrams, I/O measurements, and time course at f2 ¼ 9600 Hz Parameter

DPgram or I/O

Time Course (at f2 ¼ 9600 Hz)

Sampling frequency

131,079 Hz

32,765 Hz

Bandwidth

0–51,200 Hz

0–12,800 Hz

Number of lines

6400

800

Frequency resolution

8 Hz

16 Hz

Duration of an elementary block

125 ms

62.5 ms

Number of means

4

22

Type of average

Linear

Linear

Duration for one recording

250 ms

500 ms

l

If the difference between the DPOAE and the background noise is less than 3 dB, the emergence of the DPOAE signal compared to the noise is considered insufficient; as a result, the spectrum is not used for the calculation of the average DPOAE level. This difference is calculated by subtracting to the DPOAE value obtained on the FFT line “n” (corresponding to 2*f1
f2), the background noise calculated by averaging eight close FFT lines (n
6 to n
3 and n + 3 to n + 6). If 75% or more of the instantaneous spectra are rejected, the DPOAE application

102

Thomas Venet et al.

indicates that the entire acquisition is unusable for determining the average DPOAE value reliably. Spectra averaged over 250 or 500 ms are stored in memory. They can be plotted in waterfall or time / frequency diagrams. One frequency can also be extracted and displayed in the form of a two-dimensional time/intensity graph. During DPgram or I/O measurements, the average DPOAE level is determined by the linear average of 20 average spectra of 250 ms (i.e., 80 instantaneous spectra), which corresponds to a 5-s acquisition time (excluding rejection). This duration is long enough to obtain a reliable average value. 3.3 DPOAE Analysis and Interpretation 3.3.1 Input/Output (I/O) Measurement

3.3.2 DPgram Measurement

The I/O measurements allow the response of the cochlea at both low and medium intensities to be explored. This type of measurement provides a large amount of information, but requires a significant acquisition time per frequency, which has for consequence a limitation of the number of frequencies that can be measured. The example below shows how the choice of the level of stimulation can modify the measured effect and affect the interpretation. Figure 7 shows the average I/O DPOAEs obtained for 17 untreated Brown Norway rats. These animals were measured at 6 and 24 months. The variations recorded correspond to the effect of aging on hearing performances called presbycusis. Presbycusis includes several mechanisms, including a loss of hair cells in apical and basal extremities of the organ of Corti. These results clearly illustrate that the measured age effect depends on the intensity of the primaries L1 and L2. The age-related DPOAE decrease would be of 3.2 dB for a measurement at L1 ¼ 50 dB, whereas the difference would be close to zero with L1 ¼ 70 dB. In this experiment, L1 ¼ 40 dB was the lowest level usable due to the proximity of the background noise, notably for the 24-month measurements (see Note 9). From our experience, DPOAEs measurements are more sensitive to damage to OHCs for L1 intensities below 60 dB. Therefore, DPgram measurements (see below) should preferably be carried out between 50 and 60 dB in order to obtain a good compromise between sensitivity to cochlear damage, the intensity of the measured signal, and the emergence from the noise floor. Performing a DPgram consists in recording DPOAEs at different f2 frequencies with the same level of stimulation. This procedure is commonly called DPgram because, graphically, it resembles a clinical audiogram. The DPgram is very convenient because it allows the exploration of a wide frequency range, and therefore a large portion of the organ of Corti, in a minimum amount of time. Figure 8 presents different ways to display DPgrams obtained in control and exposed Brown Norway rats [16]. The treatment

How to Measure Distortion Product Otoacoustic Emissions in Rats?

103

Fig. 7 DPOAE levels measured in a non-exposed group (n ¼ 17) of Brown Norway rats at 6 and 24 months of age. f2 ¼ 25,440 Hz with f2/f1 ¼ 1.2 and L1 ¼ 40–75 dB with L1 ¼ L2 + 11 dB. The error bars represent the standard deviation

was a co-exposure to impulsive noise and to 600-ppm styrene (6 h/ day, 5 days/week, for 4 weeks). The impulsive noise was an 8-kHz octave band burst lasting 10 ms at 112 dB, which was repeated every 15 s. This noise exposure was designed to ensure that each noise burst was absorbed by the cochlea without being reduced by the middle ear reflex. In terms of acoustic energy, the equivalent noise exposure was an 80 dB continuous noise during 8 h (LEX,8h ¼ 80 dB). The DPgram (Fig. 8a, b) was performed with L1 ¼ 55 dB, which yields DPOAE levels between 15 and 25 dB SPL depending on the frequency considered. These levels are largely sufficient not to be disrupted by the noise floor, which is comprised between
6 and
1 dB SPL at high and medium frequencies, respectively. They are also sufficient to evaluate the temporary and/or permanent impact of exposure on the cochlea that will result in a decrease in DPOAE amplitudes. In the rat, the DPOAE magnitudes are relatively low at 3600 Hz (15 dB), compared to those obtained at higher frequencies. This occurs because the lower frequency limit measurable with otoacoustic emissions (about 3000 Hz) is close. It might also be convenient to plot the DPOAEs as differences from baseline data (Fig. 8c, d). The dashed line at the “0” on the ordinate indicates no change from baseline levels. One might notice that unexposed animals display a slight shift in DPOAE levels between 3600 and 9600 Hz ( f2). This is likely due to aging (see Note 10). To take into account this drift in DPOAE amplitudes

104

Thomas Venet et al.

Fig. 8 DPgrams (a, b) and DPOAE variations (c, d) obtained in control rats (b, d) and rats exposed to impulse noise (octave band noise centered at 8 kHz, LEX,8h ¼ 80 dB) and 600-ppm styrene for 4 weeks (6 h/day, 5 days/week) (a, c). Exposure was followed by a 4-week recovery period. DPOAE measurements were performed with L1 ¼ 55 dB before exposure (baseline) and 1 day (1 day post) and 4 weeks post-exposure (4 weeks post). DPgrams (a, b) display the raw DPOAE values at different time-points for a given experimental group. DPOAE variations (c, d) are defined as [post-exposure DPOAE
baseline] for a given experimental group. DPOAE variations normalized by the controls (e) are calculated as follows: [(post-exposure DPOAEexposed
post-exposure DPOAEcontrols)
(baseline DPOAEexposed
baseline DPOAEcontrols)] and take into account the drift in DPOAE amplitudes observed in the control group

observed in the control group, one might consider normalizing the DPOAE value of the exposed animals by the values obtained in control animals (Fig. 8e) by doing the following calculation:
½ post
exposure DPOAEexposed
post
exposure DPOAE controls

baseline DPOAEexposed
baseline DPOAEcontrols : The present analysis highlights the wide frequency range affected by the noise and styrene co-exposure although the

How to Measure Distortion Product Otoacoustic Emissions in Rats?

105

impulsive noise was filtered band-pass filtered between 5600 and 11,200 Hz (octave band noise 8 kHz). A clear permanent impairment was obtained at all tested f2 frequencies but the highest, 25,440 Hz. Moreover, this DPgram analysis shows that the post-exposure recovery is highly dependent on the frequency considered (full at 25,440 Hz, limited at 9600 Hz).

4

Conclusion DPOAE measurements are particularly attractive for assessing cochlear function in laboratory animals such as rats because they are objective, robust, easily measured and offer the possibility to test a wide range of frequencies within the species’ hearing range. In addition, measuring DPOAEs in rats is fast, noninvasive, and can be performed repeatedly in the same subjects. The downside of DPOAE measurements in rats is the anesthesia, which might introduce in some cases an additional variable in the experiment. It is also of importance to keep in mind that, in rats, frequencies below 3000 Hz cannot be investigated using this method, and that it is necessary to maintain steady body temperature and atmospheric pressure. Despite these limitations, DPOAEs are extremely valuable because OHCs are very sensitive to chemical impairment, noise and oxidative stress, among other nuisances. For a thorough investigation of the health of the cochlea and central auditory pathways, additional auditory measurements can be performed, such as cochlear microphonic response and auditory brainstem-evoked responses (ABR). The cross analysis of these two measurements can help to localize the origin of the impairment. DPOAEs are specific to OHCs and can be measured even if IHCs are damaged [9, 17]. In contrast to DPOAEs, ABR thresholds are dependent on the function of the whole auditory system, i.e., OHCs, IHCs, and ascending auditory pathways and nuclei [18]. In addition, depending on your scientific question, you might consider complementing these functional techniques by cochlear histology, such as cytocochleograms (see Chap. 2), scanning electron microscopy of the stereocilia bundles, or mid-modiolar sections of resin-embedded cochleae.

5

Notes 1. At the time this text is being written, two companies (TuckerDavis Technologies and Intelligent Hearing systems) provide integrated commercial systems allowing DPOAE measurements up to 32–40 kHz in rodents. Although these systems have not been tested by the authors, they have been previously

106

Thomas Venet et al.

used in several published studies [14, 19] and should be considered as viable and simple alternatives to our custom setup. 2. Ideally, DPOAE measurements should be performed within a double-walled sound-attenuation chamber to reduce background noise. The chamber should include a small double-pane window or a video camera for observing the animal during testing and should be large enough to stand inside it. In case such an equipment is not an option, you might consider a single-walled chamber, or tabletop sound booth. However, always attempt to maintain the noise floor as low as possible by eliminating/concealing all surrounding noise sources. 3. Using two different generators, each connected to a transmitter, reduces the risk of acoustic distortion generated by the equipment setup itself. Indeed, electrical and acoustic signal distortions easily occur in a measurement system. Such signal interferences, mixed to the physiological signal, might render the recording unusable. It is very important to ascertain that the signal is purely of physiological origin. 4. All the elements composing our prototype probe are glued together with cyanoacrylate glue so that the manipulation of the probe does not create any air-leak, which could modify sound-wave impedance. Electrical ground loops must be eradicated and shield must be thorough. 5. As an indication, the uncertainty widened to a level of confidence of 95% (widening factor 2.26) is 0.5 dB at 4000 Hz and 1.0 dB at 16000 Hz for the levels L1 and L2 of the primaries f1 and f2. The uncertainty increased to a 95% confidence level (2.26 enlargement factor) of the sound pressure measured by the probe’s microphone is 0.9 dB at 4000 Hz and 1.8 dB at 16000 Hz. 6. When anesthetized, the body temperature of the rats drops rapidly. It is very important to maintain it stable using a heating blanket with a feedback probe (38  C). If the rat gets cold, DPOAE responses are profoundly reduced and become unusable. This occurs because the eardrum gets “depressed.” A “depressed” eardrum is easily visible by otoscopy as it gets very close to the ossicles and let them appear by transparency. In this case, warm up the rat for a few minutes and check that eardrum is slightly “inflated” towards you. 7. You should ascertain that there is no difference of pressure between the room where the animal has been anesthetized and the measurement booth. The anesthetized rat does not regulate the balance of pressures on either side of the eardrum. Performing a cochlear paracentesis [20] can solve the problem for a single measure of DPOAE, but is not recommended for

How to Measure Distortion Product Otoacoustic Emissions in Rats?

107

repeated measurements because of the risks of eardrum damage and infection. 8. When stimulated by a two-tone stimulus, the organ of Corti produces a family of distortion products having specific arithmetic relationships with the frequencies of the stimulus tones (3*f1
2*f2, 4*f1
3*f2; 2*f2
f1; 3*f2
2*f1, 4*f2
3f1; f2
f1, etc.). However, the most robust and easily recorded in both laboratory animals and human is the DPOAE at 2*f1
f2. 9. With L1 < 40 dB (data not shown), the difference in DPOAE levels between 6 and 24 months decreases because DPOAE levels hardly emerge from background noise. Only rats with the highest DPOAEs can be measured with such a low L1, especially at 24 months. Therefore, the data are no longer representative of the group but only of the rats having the best hearing performances. 10. During an experimental protocol of DPOAE measurements with rats, it is important to carry out the measurements of a control group simultaneously with the exposed group. The two groups must be measured under the same conditions in order to be able to analyze any variations and thus reinforce the accuracy of these measurements. References 1. Mahendrasingam S, Beurg M, Fettiplace R, Hackney CM (2010) The ultrastructural distribution of prestin in outer hair cells: a postembedding immunogold investigation of low-frequency and high-frequency regions of the rat cochlea. Eur J Neurosci 31 (9):1595–1605. https://doi.org/10.1111/j. 1460-9568.2010.07182.x 2. Fettiplace R (2017) Hair cell transduction, tuning, and synaptic transmission in the mammalian cochlea. Compr Physiol 7 (4):1197–1227. https://doi.org/10.1002/ cphy.c160049 3. Jaramillo F, Markin VS, Hudspeth AJ (1993) Auditory illusions and the single hair cell. Nature 364(6437):527–529. https://doi. org/10.1038/364527a0 4. Strelioff D, Flock A (1984) Stiffness of sensory-cell hair bundles in the isolated Guinea pig cochlea. Hear Res 15(1):19–28. https:// doi.org/10.1016/0378-5955(84)90221-1 5. Khanna SM (2002) Non-linear response to amplitude-modulated waves in the apical turn of the Guinea pig cochlea. Hear Res 174 (1–2):107–123. https://doi.org/10.1016/ s0378-5955(02)00645-7

6. Howard J, Hudspeth AJ (1988) Compliance of the hair bundle associated with gating of mechanoelectrical transduction channels in the bullfrog’s saccular hair cell. Neuron 1 (3):189–199. https://doi.org/10.1016/ 0896-6273(88)90139-0 7. Kemp DT (2002) Otoacoustic emissions, their origin in cochlear function, and use. Br Med Bull 63:223–241. https://doi.org/10.1093/ bmb/63.1.223 8. Subramaniam M, Salvi R, Spongr V, Henderson D, Powers N (1994) Changes in distortion product otoacoustic emissions and outer hair cells following interrupted noise exposures. Hear Res 74(1–2):204–216 9. Trautwein P, Hofstetter P, Wang J, Salvi R, Nostrant A (1996) Selective inner hair cell loss does not alter distortion product otoacoustic emissions. Hear Res 96(1–2):71–82 10. Hofstetter P, Ding D, Powers N, Salvi RJ (1997) Quantitative relationship of carboplatin dose to magnitude of inner and outer hair cell loss and the reduction in distortion product otoacoustic emission amplitude in chinchillas. Hear Res 112(1–2):199–215

108

Thomas Venet et al.

11. Emmerich E, Richter F, Reinhold U, Linss V, Linss W (2000) Effects of industrial noise exposure on distortion product otoacoustic emissions (DPOAEs) and hair cell loss of the cochlea–long term experiments in awake Guinea pigs. Hear Res 148(1–2):9–17 12. Shera CA (2004) Mechanisms of mammalian otoacoustic emission and their implications for the clinical utility of otoacoustic emissions. Ear Hear 25(2):86–97. https://doi.org/10.1097/ 01.aud.0000121200.90211.83 13. Martin GK, Stagner BB, Lonsbury-Martin BL (2006) Assessment of cochlear function in mice: distortion-product otoacoustic emissions. Curr Protoc Neurosci. Chapter 8: Unit8.21C-Unit28.21C. https://doi.org/10. 1002/0471142301.ns0821cs34 14. Sheppard AM, Zhao DL, Salvi R (2018) Isoflurane anesthesia suppresses distortion product otoacoustic emissions in rats. J Otol 13 (2):59–64. https://doi.org/10.1016/j.joto. 2018.03.002 15. Robles L, Ruggero MA (2001) Mechanics of the mammalian cochlea. Physiol Rev 81 (3):1305–1352. https://doi.org/10.1152/ physrev.2001.81.3.1305

16. Venet T, Campo P, Thomas A, Cour C, Rieger B, Cosnier F (2015) The tonotopicity of styrene-induced hearing loss depends on the associated noise spectrum. Neurotoxicol Teratol 48:56–63. https://doi.org/10.1016/j.ntt. 2015.02.003 17. Le Calvez S, Avan P, Gilain L, Romand R (1998) CD1 hearing-impaired mice. I: distortion product otoacoustic emission levels, cochlear function and morphology. Hear Res 120(1–2):37–50. https://doi.org/10.1016/ s0378-5955(98)00050-1 18. Kujawa SG, Liberman MC (2009) Adding insult to injury: cochlear nerve degeneration after “temporary” noise-induced hearing loss. J Neurosci 29(45):14077–14085. https://doi. org/10.1523/JNEUROSCI.2845-09.2009 19. Dong W, Stomackin G, Lin X, Martin GK, Jung TT (2019) Distortion product otoacoustic emissions: sensitive measures of tympanicmembrane perforation and healing processes in a gerbil model. Hear Res 378:3–12. https:// doi.org/10.1016/j.heares.2019.01.015 20. Bergin M (2013) Systematic review of animal models of middle ear surgery. World J Otorhinolaryngol 3(3). https://doi.org/10.5319/ wjo.v3.i3.71

Chapter 6 Use of Visual Evoked Potentials to Assess Deficits in Contrast Sensitivity in Rats Following Neurotoxicant Exposures William K. Boyes Abstract This chapter describes a procedure for recording pattern-elicited visual evoked potentials from experimental animals, focused primarily on pigmented rats. When recorded over a range of visual pattern contrast values, the results can be used to derive estimates of visual contrast threshold, contrast sensitivity, and contrast gain. Visual contrast is defined as the difference between the bright and dark regions of a visual pattern, adjusted for the overall luminance. Contrast encoding is an important feature of the neurological processes underlying spatial vision and is dependent on integrated processing within defined neurological circuits. This chapter describes procedures to measure contrast-related parameters that have been developed over years of experience and trial and error approaches. They involve electrophysiological recordings from visual cortex while animals view modulating visual patterns. The resulting evoked potentials are signal averaged, subjected to spectral analysis and interpreted relative to the contrast of the eliciting visual patterns. The resulting parameters include measurements of response amplitude, contrast threshold, contrast sensitivity, and contrast gain. The data from experimental animals are highly analogous to those from human subjects and have shown similar responsivity to neurotoxicant exposures. Key words Visual evoked potentials, Contrast threshold, Contrast sensitivity, Contrast gain, Developmental neurotoxicity

1

Introduction Testing visual function for potential toxic effects is important as a primary adverse outcome and is also an approach to assess neurological function. The retina is part of the central nervous system and projects to the visual centers of the brain. The visual system is accessible for well-controlled stimulation that enables careful assessments of many aspects of neurological function. It is important to distinguish optical from neurological aspects of vision. A typical examination of visual acuity, in which a patient observes black letters on a white background and reports the letters of the smallest font that are discernable, is primarily an evaluation of the

Jordi Llorens and Marta Barenys (eds.), Experimental Neurotoxicology Methods, Neuromethods, vol. 172, https://doi.org/10.1007/978-1-0716-1637-6_6, © Springer Science+Business Media, LLC, part of Springer Nature 2021

109

110

William K. Boyes

optical apparatus (cornea and lens) at the front of the eye. Visual acuity measurement depends on the ability to make a well-focused image on the retina and works well for determining refraction values for glasses or contact lenses. However, a measure of visual acuity does not evaluate much of the neural aspects of visual function. Visual perception is important over a range of visual pattern sizes, not just the smallest letters identifiable. Vision also requires encoding features of the visual world such as pattern size, shape, motion, color, and depth. One of the most important features of visual pattern perception is contrast, expressed as the luminance difference between the bright and dark parts of a visual pattern, adjusted for the overall luminance. Contrast thresholds are defined conceptually as the lowest contrast level which is readily visible. In psychophysical experiments, contrast threshold might be defined by the lowest contrast target that is detected statistically on half or more of the trials presented. Using visual evoked potentials (VEPs), (see Note 1), contrast thresholds can be thought of as the lowest contrast stimulus to elicit a response with an amplitude above zero, or alternatively, above the recording noise level. Visual contrast sensitivity, the inverse of contrast threshold, turns out to be highly sensitive to disruption by exposure to a range of neurotoxic compounds. Prior to describing a procedure for testing visual contrast perception in experimental animals, a few basic concepts relating to the neurological processes encoding spatial and temporal vision will be considered. Figure 1 presents a simplified and stylized scheme of the retina. The photoreceptors (rods and cones; P) lie at the back of the retina with cell bodies in the outer nuclear layer. The photoreceptors receive photons of light and generate neurological signals that feed directly into retinal bipolar cells (B), which in turn project to retinal ganglion cells (G). The retinal ganglion cells send axons via the optic nerve to the visual areas of the brain. At the retinal outer plexiform layer, where the photoreceptors meet the bipolar cells, there are also synaptic contacts with horizontal cells (H). The horizontal cells form inhibitory connections across neighboring photoreceptor-bipolar cell channels, so that when one channel is activated by light, the neighboring channels are correspondingly deactivated, and vice versa. This has the effect of giving bipolar cells and retinal ganglion cells what are referred to as “center-surround” receptive fields (Fig. 2). The receptive field of a neuron is that portion of the receptor surface (in this case the retina) which, when stimulated, causes the neuron to increase or decrease its firing rate. Light falling on the center of the receptive field might activate a cell, while stimulation of the surrounding area would inhibit it, forming a center-surround topography. The retina contains cells with both on-center and off-center receptive fields. For off-center cells, stimulation of the center inhibits, and the surround activates, the cell. The receptive fields also vary in size, with small receptive

Visual Contrast Sensitivity Testing

-

- -

H

Inner nuclear layer Horizontal cells (H) Bipolar cells (B) Amacrine cells (A)

B

P

-

-

Outer plexiform layer

Inner plexiform layer

P

P

Outer nuclear layer Photoreceptors (P)

111

H B

A

B A

Off On

Ganglion cell layer Retinal ganglion cells (G) Optic nerve

G

G

G

Light

Fig. 1 A schematic representation of retinal connectivity. The three photoreceptors (P) are depicted with their cell bodies located in the outer nuclear layer. The photoreceptors make synaptic connections in the outer plexiform layer with retinal bipolar cells (B) and horizontal cells (H). Horizontal cells make inhibitory contact between the photoreceptor/bipolar cell interface so that when a photoreceptor and bipolar cell become activated the neighboring bipolar cells are inhibited. Conversely, when one of the photoreceptors depicted on the side is activated, the bipolar cell in the middle is inhibited. The bipolar cells which in turn contact retinal ganglion cells (G) and amacrine cells (A) in the inner plexiform layer. Synaptic connections in the outer portion of the inner plexiform layer primarily signal stimulus offsets, whereas synaptic contacts in the inner portion of the outer plexiform layer primarily signal stimulus onsets. The axons of retinal ganglion cells form the optic nerve, leave the eye, and project to visual areas of the brain. With this simplified diagrammatic arrangement, the primary spatial and temporal features of visual pattern perception are established

fields tuned for fine patterns and large receptive fields for large patterns. The size of visual patterns can be measured as spatial frequency in units of the cycles of a repetitive pattern per degree of visual angle (cpd). The luminance difference between the center and surround portions of the receptive field governs the neuronal firing rate, and hence cellular sensitivity to contrast. Therefore, in the combination of spatial frequency and contrast, the basic elements of perceiving visual patterns are encoded in the outer plexiform layer of the retina. Although sensitivity to contrast begins in the retina, it is tuned and refined at later stages of the visual system.

112

William K. Boyes

P

P

P

Stimulus Luminance

on

Sustained Response

Transient Response

off

H B

H B

A

B A

Off On

G

G

G

Light

-

+

-

Distance

Time

Time

Fig. 2 Left Column. The retinal cellular connectivity from Fig. 1 is reproduced along with a depiction of an excitatory center-surround receptive field. Receptive fields reflect region of the innervated surface (in this case the retina) for which the neuronal firing rate changes (increase or decrease) with stimulation. The horizontal cell inhibitory interactions with neighboring photoreceptor/bipolar cell connections has the result of producing center receptive fields as shown below. Therefore, if the neural activity pattern of a retinal ganglion cell is measured it has a concentric “center-surround” spatial organization as depicted from the retinal surface perspective in the circles below. Stimulation in the central area of the receptive field causes the firing rate of this retinal ganglion cell to increase, while stimulation in the surround region causes the firing rate to decrease. The size of the receptive field reflects the extent of the reach of horizontal cell processes across the retinal surface. The amount of increase or decrease in firing rate is a function of the total luminance difference between the center and surround regions Second column from left: the three-part panels represent the stimulus luminance impinging on the center or surround regions of the receptive field shown at bottom left. The four panels at right depict the neuronal firing rate recorded in response to the associated stimulation Top: a high-contrast stimulus with the bright part over the excitatory region and the dark part over the inhibitory region causes the neuron to display a maximum firing rate Second panel: the contrast between the bright and dark parts of the stimulus is reduced Third panel: a high-contrast stimulus in reverse phase to the stimulus in the top row Bottom: a lower contrast, reverse-phase stimulus Third column from left shows hypothetical “sustained” firing rate patterns elicited by stimuli on the same row at the left. Top: The neuron shows a sustained increase in firing rate while the stimulus is presented. Second panel: firing rate proportionally less than maximal because of the lower contrast stimulus. Third panel: the high-contrast stimulus is reversed. The reverse-phase stimulus causes a reduction of firing rate that is proportional to the summed stimulation of the center and surround portions of the receptive field. Bottom. Reverse-phase lower contrast stimulus causes a lesser degree of inhibition of the normal firing rate Right column. Hypothetical responses of cells firing in a “transient” manner to the stimuli shown in second column. Top: Sustained stimulation with high-contrast stimuli causes robust responding when turned on or off, which fades after the stimulus transition. Second panel: Transitions of lower contrast stimulus cause transient bursting responses of lesser intensity than the higher contrast stimulus. Third panel: Reverse-phase transitions cause transient responses similar to top row. Low--contrast stimulus transitions cause response firing patterns similar to the second row

Visual Contrast Sensitivity Testing

113

The retina also encodes important temporal aspects of visual stimuli, which happens primarily at the inner plexiform layer. Temporal tuning involves synaptic activity of amacrine cells (A) as they influence transmission between bipolar cells and retinal ganglion cells. Off-signals are transmitted in the outer portion of the inner plexiform layer, while on-signals are processed in a distinct layer in the inner portion of the inner plexiform layer. Electrophysiological recordings from retinal ganglion cells reveal two distinct temporal profiles of neuron activity (Fig. 2). In response to light falling on their receptive fields, some neurons show a steady firing rate, either an increase for on-center cells or a decrease for off-center cells. These are sometimes referred to as “sustained” responses. Other cells show bursts of firing whenever the light is either turned on or turned off, but the change in firing rate rapidly adapts if the stimulation is prolonged. These are referred to as “transient” responses. Some think of sustained responders as “pattern detectors” since they respond whenever a pattern is present, and transient responders as “motion detectors” since they respond whenever a pattern is changing. Another terminology for sustained and transient response patterns is “linear,” and “nonlinear,” respectively, relating to the fidelity in which neuronal response rates follow temporal and spatial aspects of visual pattern modulation. There are variations in the proportions of cells with linear or nonlinear responses among species. Human, non-human primates, and many mammalian retinas have both sustained and transient responding neurons. Rats tend to have a preponderance of transient, “motion-detector” neurons, and under steady-state sinusoidal modulating patterns, the rat visual system responds primarily as a frequency-double (“nonlinear”) rate, reflecting bursts of responses to both the on and off cycles of a temporally modulated visual pattern (Fig. 3). This chapter describes procedures to record VEPs as a method to assess visual pattern perception and sensitivity to visual contrast. These procedures were originally adopted from human clinical and psychophysical practices and have proved to be sensitive measures of potential neurotoxicity. Literature on the basic theory and practice of electrophysiological recordings, including sensory evoked potentials and their use in neurotoxicology, are available elsewhere (1–5). Rats were selected as the primary subjects for study because, at the time these procedures were developed, the rat was a primary subject of experimental neurotoxicology research. Sensory evoked potential procedures also work in mice (6, 7) and other species such as cat, pig, or primate (8–10). It is important, however, that a pigmented strain of rat or mouse is used (see Note 2), because albino strains of rodents, lacking melanin in the retinal pigment epithelium and other ocular structures, have poor pattern vision and give inferior pattern VEPs (11). The procedures involve surgically implanting indwelling recording electrodes in the skull located

114

William K. Boyes

Temporal Sine Wave

Temporal Square Wave 1.5 1.0 0.5

on

0.0 -0.5

Stimulus

-1.0 -1.5 0.00

off

0.05

0.10

0.15

0.20

Time (sec) 1.5

F1 Waveform, 100 % Contrast

Sustained Response

Contrast

1.0 0.5 0.0 -0.5 -1.0 -1.5

0.0 1.5

0.5

1.0

1.5

2.0

2F Waveform, 100 % Contrast

Transient Response

Contrast

1.0 0.5 0.0 -0.5 -1.0 -1.5 1.5

0.0

0.5

Contrast

Transient Response 50% contrast

1.0

1.5

2.0

F2 Waveform, 50 % Contrast

1.0 0.5 0.0 -0.5 -1.0 -1.5

0.0

0.5

1.0

1.5

2.0

Time (sec)

Fig. 3 Illustration of how hypothetical sustained and transient responses translate into steady-state VEPs at the stimulus rate (F1) and twice the stimulus rate (F2), respectively. Basically, the firing rate of sustained neurons follow the temporal modulation of the pattern (F1), while the transient neurons, responding to both on and off phases of the stimulus transitions, respond at twice the stimulation rate (F2). Stimulation of highcontrast patterns causes a strong F2 evoked potential (third row). Stimulation with a pattern 50% reduced in contrast causes a lower amplitude F2 response (bottom row)

over visual cortex (see Note 3) (Fig. 4), followed by approximately 1 week for surgical recovery prior to VEP recording. The recordings are performed using awake rats (Fig. 5) in order to avoid the effects of anesthesia on the cortical VEPs and potential interactions between the anesthetics and the actions of potential neurotoxic substances being studied. The awake animals are mildly restrained (see Notes 11, 14, 16) during testing to assure that they are watching the visual stimuli and are at a constant distance (15 cm) and orientation to the stimulus screen. Computer-generated stimulus patterns are presented on a video monitor for the subject to observe, modulated over time, and electrical activity from visual cortex is concurrently recorded. The electrophysiological activity is amplified, filtered, and artifacts from movement and other causes are detected and removed. The cleaned electrophysiological data are signal averaged in synchrony with the temporal modulation of the visual pattern in order to enhance neuronal activity related to

Visual Contrast Sensitivity Testing

115

Connector and cable leading to amplifier

Cap / connector

Assembled electrode

Electrode components: Pin Wire Screw

Fig. 4 Photograph of components for construction of cranial screw electrodes and connection to amplifier input cables. Lower left: assembled electrode. Lower right: electrode components including: stainless-steel screw, insulated wire with insulation scrapped from both ends for soldering to screw, and gold pin to crimp on the top of electrode wire. The wire is soldered to the screw leaving the screw slot opened. The screw component is threaded to a matched size hole in the skull over visual cortex to provide the recording surface. The pin is placed into a designated hole in the cap/connector (an Amphenol 9 pin connector) for top of headset. The headset assembly (also including ground and reference electrodes not pictured) is cemented in place. After surgical recovery, the cap is temporarily attached to the matched connector and cable leading to the amplifier for a recording session

visual stimulation (“signal”) and reduce non-synchronized activity which presumably is unrelated to visual signal processing (“noise”). The resulting VEP is submitted to post hoc analyses, in our case Fast Fourier Transformation (FFT) processing, that yields the signal amplitude at the stimulus rate and harmonic frequencies. Stimulus amplitude data are then expressed as a function of the visual stimulus contrast to determine contrast sensitivity and contrast gain parameters. Response amplitudes, contrast sensitivity, and contrast gain values can be compared across groups of subjects that were treated with vehicle (i.e., control) or different doses of the test material. The experimenter has multiple options regarding temporal and spatial parameters of the visual stimuli, data recording, and analyses. We will discuss those parameters and the rationale for selections below.

116

William K. Boyes

Fig. 5 Schematic illustration of procedure for recording pattern-elicited VEPs from awake rats (reproduced with permission from Boyes, 1994) 1.1 Stimulus Temporal Parameters

Several different modes of temporal modulation of the visual patterns have been used. The preferred option in our laboratory is steady-state pattern-reversal modulation. The steady-state term implies that the stimulus modulation and the resulting evoked potentials are continuous, in contrast to “transient” evoked potentials, in which the stimuli change abruptly and discrete responses to each stimulus presentation are observed. With transient stimulation, the evoked potential waveforms are typically complex, with multiple positive and negative peaks, and are usually scored as the latency and amplitude of each positive and negative going peak. In some cases, the underlying neurological generators of these peaks have been identified which can be an advantage for interpretation of results. In other cases, however, the peak generators are unknown or disputed, and neurotoxicological treatments may alter the waveform shape so that the identification of individual peaks becomes problematic. In addition, scoring the entire waveform with multiple peaks and valleys produces a large number of dependent variables (twice the number of peaks and troughs if the latency and amplitude of each is measured). Statistically, recording multiple dependent measures from individual subjects causes an increased probability of incorrectly rejecting the null hypothesis, which should be addressed through a correction of the alpha level using appropriate statistical adjustments. In steady-state recordings, in contrast, the stimulus modulation is continuous, and the stimulus rate is adjusted so that the response from one transition blends into another and the evoked potential takes on a sinusoidal characteristic (11). In this case, it is possible to submit the waveform to a Fourier analysis and obtain amplitude of the primary Fourier component(s) of the waveform. This is an unambiguous measure of the strength of the response that is not dependent on the semi-subjective selection of peaks for scoring.

Visual Contrast Sensitivity Testing

117

1.5

Contrast

1.0

Temporal modulation (1F)

0.5 0.0 -0.5 -1.0 -1.5 0.0

0.5

1.0

1.5

2.0

1.5

2.0

1.5

2.0

Time (sec) 1.5

Contrast

1.0

Hypothetical transient response (2F)

0.5 0.0 -0.5 -1.0 -1.5 0.0

0.5

1.0

Time (sec) 1.5

Contrast

1.0

Hypothetical transient response, (2F); 50% contrast

0.5 0.0 -0.5 -1.0 -1.5 0.0

0.5

1.0

Time (sec)

Fig. 6 Left. Sine waves at the stimulus rate (Top), a high-contrast and twice the stimulus rate (second row) and at 50% contrast, and twice the stimulus rate (Bottom row) Center: group averaged VEP waveforms from rats in response to stimuli of varying contrast as noted. Right column. Spectral analysis of the waveforms to the left showing predominant F2 amplitude peak, which increases as a function of the stimulus contrast. Higher frequency harmonic responses are also evident in the spectral analysis of the higher contrast responses

The rat visual system responds primarily in a transient mode—rats have much more visual neurons that respond in a transient than a sustained manner, and the rat waveform shows a strong frequency—double (nonlinear) response component labeled F2. The amplitude of the F2 component therefore becomes an unambiguous, objectively scored, single response measure (Fig. 6). In addition to transient and steady-state presentation, there is the choice of pattern-reversal (“counterphase”) or pattern on-off (appearance–disappearance) modulation (Fig. 7). For patternreversal modulation, the dark and light portions of the stimulation alternate fully in a counterphase fashion so that every part of the screen changes over time from dark to light and back again. In on-off, the pattern modulates from a mean luminance screen to one phase of the spatial stimulus and back to a mean luminance screen. On-off has an advantage in that steady-state responses to on-off modulation contain both F1 (“linear”) and F2 (“nonlinear”) components (12). Pattern-reversal stimulation reveals only F2 (“nonlinear”) responses in the VEP waveform. Counterphase has the advantage that the contrast modulation is effectively twice that of on-off modulation, yielding stronger, and more robust evoked potentials. The F1 (linear) responses of rats are very small in comparison to humans, which reflects the high proportion of the visual

118

William K. Boyes

1.5

Pattern Reversal

1.0

1.0

0.5

0.5

Contrast

Contrast

1.5

0.0

On / Off

0.0

-0.5

-0.5

-1.0

-1.0 -1.5

-1.5 0.0

0.1

0.2

0.3

Time (sec)

0.4

0.5

0.0

0.1

0.2

0.3

0.4

0.5

Time (sec)

Fig. 7 An illustration of pattern-reversal (“counterphase) vs On/Off (“appearance–disappearance”) modulation of the visual patterns. In pattern-reversal (left side), the stimulus pattern alternates so that each portion of the stimulus reverses from light to dark phase of the pattern and back again with each temporal cycle. In On/Off modulation (right side), the stimulus transitions between a mean luminance, zero contrast pattern, and the full pattern and back on each temporal cycle

neurons of the rat responding in a transient fashion. In practice, we have never seen a robust toxic effect in the rat F1 response. Therefore, we now prefer using pattern-reversal modulation for most testing applications due to the F2 responses being larger, less variable and having a higher signal-to-noise ratio. 1.2 Stimulus Spatial Parameters

Spatial Frequency. Spatial frequency refers to the size of a repetitive visual grating and is expressed as cycles of the pattern/degree visual angle (cpd). Human subjects are typically tested over a range of spatial frequencies. The range of spatial frequency sensitivity in rats, however, is more limited. We use stimuli of about 1.6 cpd, which is the approximate peak of the contrast sensitivity function of pigmented rats (13). Contrast. Contrast (C) is defined as the difference between the maximum and minimum luminance (Lmax and Lmin, respectively) of a visual pattern, adjusted for the overall background luminance. C ¼ ðLmax  Lmin Þ=ðLmax þ Lmin Þ Contrast values range from 0 (no pattern) to 1 (all the light from the bright parts of a pattern). Frequently, contrast values are multiplied by 100 to convert them into percent contrast (ranging from 0 to 100%). The F2 amplitude of the rat steady-state pattern-reversal VEP shows a predictable relationship to the contrast of the visual stimulus. The F2 amplitude increases in a linear fashion with the log of the stimulus contrast (14, 15). This relationship follows Weber-

Visual Contrast Sensitivity Testing

119

Fechner law of sensory psychophysics, which contends that the perceived intensity of sensory stimuli is a linear function of the log stimulus intensity. The perceived intensity of sensory stimulation is encoded by the firing rate of sensory neurons. Therefore, the log-contrast F2 amplitude functions of pattern VEPs reflect original coding of differential stimulation of the center-surround receptive fields of retinal neurons, translated into the firing rates of visual neurons, and carried on ascending sensory pathways to the visual areas of the brain. Another concept is contrast gain, which refers to the increased neuronal firing rate with increased contrast of visual patterns. Contrast gain adjustment refers to the adaptation of the firing rate of sensory neurons in the face of altering visual contrasts. Adaptation enables the sensitivity to be adjusted so that subtle stimuli can be perceived in a low-contrast environment, such as a dim-light overcast day, and also not overwhelmed on a bright sunny high-contrast day. The firing rates of visual neurons are adjusted to changing visual contrast to maintain an optimal firing range. Contrast gain adjustments of visual neurons have been observed to some extent in neurons of the retina and lateral geniculate nucleus of the thalamus, but the primary site of contrast gain adjustment is thought to be neurons of visual cortex (16). In VEPs, the slope of the amplitude log-contrast functions has been interpreted to reflect contrast gain (15). VEP log-contrast amplitude functions therefore reflect the sensitivity of the visual system to contrast as well as the adjustments of the system to changing contrast levels. We recently reported that adult rats, who were deficient in thyroid hormone during their pre- and postnatal development due to perinatal exposure to propylthiouracil (PTU), had doserelated decreases in the slope of pattern-reversal VEP log-contrast F2 amplitude functions (Fig. 8) (17). Propylthiouracil causes a deficit in thyroid hormones by inhibiting iodine transport in the thyroid gland. Thyroid hormones are critical for neurodevelopment, including for neurons in the visual system. There was evidence that PTU-treated rats had both retinal deficits (decreases in green flicker electroretinograms), as well as deficits in the superficial layers of visual cortex (decreases in visual contrast gain). A reduction of thyroid hormones during pre- and postnatal neurodevelopment altered the visual system in a way that persisted into adulthood, long after the thyroid hormone levels of the grown pups had returned to normal. Many environmental contaminants may alter expression of thyroid hormones, but the level of thyroid hormone inhibition and the consequences for neurodevelopment are often unclear. It is possible that assessment of contrast sensitivity and VEP contrast amplitude functions could serve as one approach to evaluate residual neurological effects of developmental thyroid disruption.

120

William K. Boyes

F2AAmmpplitliutuddee(µV) (uV)

14 12

1 ppm

0 ppm b[0] -9.2 b[1 ] 10.4 r ² 0.97

3 ppm

2 ppm b[0] -6.3 b[1] 7.07 r ² 0.92

b[0] -9.5 b[1] 10.1 r ² 0.95

b[0] -4.2 b[1] 5.2 r ² 0.98

10 8 6 4 2 0 10

100

10

100

10

100

10

100

Contrast (%) Fig. 8 The effects of perinatal dose-related inhibition of thyroid hormone synthesis on adult rat contrast vision. The time line of the experiments is depicted along the top of the figure. Pregnant rats were treated with 0, 1, 2, or 3 ppm propylthiouracil (SP) PTU in drinking water beginning on gestational day (GD) 6. The treatment continued through the duration of gestation and postnatally until postnatal day (PN) 21 when the pups were weaned and received control drinking water thereafter. PTU inhibits uptake of iodine into the thyroid gland and caused a graded amount of developmental insufficiency of thyroid hormone. The pups showed dose-related reductions of serum T4 during the treatment period, but T4 levels had returned to normal by the time of VEP testing. When the rats were adults, they were surgically implanted with recording electrodes and tested for VEPS about a week later. The four panel graphs depict the pattern VEP F2 amplitude measured from FFT analysis of the individual rat waveforms. The horizontal axes of each panel depict stimulus contrast on a repeated log scale. The vertical axis depicts F2 amplitude and is the same for each dose group. Linear functions were fit to the F2 amplitude log-contrast functions from each dose group. The slope of the functions (i.e., contrast gain) was significantly reduced following either 2 or 3 ppm PTU (Reproduced with permission from Boyes et al., 2018)

The neurons in the visual pathway operate by the same set of biochemical and neurophysiological processes as other neurons of the central nervous system. The primary excitatory and inhibitory neurotransmitters are present. They operate in sophisticated and integrated networks. Knowledge of the neural mechanisms for encoding visual pattern and contrast and the ability to precisely stimulate visual pathways with computer-generated visual patterns, make visual system testing a sophisticated and sensitive approach to evaluate potential neurotoxicity.

Visual Contrast Sensitivity Testing

121

The following sections describe protocols to record patternreversal evoked potentials from awake rats, as was reported previously (17). These procedures were built on and evolved from those used before. Interested readers may wish to find additional applications of these and related procedures (18–22).

2

Materials l

Subjects: Adult pigmented rats. They should be about 60 days of age or older at the time of surgery so that growth of the skull has slowed, and the implanted electrode assembly will remain intact. Pigmented strains are preferred because albino strains of rats have aberrant retinas and give poor pattern evoked potentials (11). Toxicology studies typically require a vehicle treated control group and one or preferably more dose groups for dose– response studies. As a general rule of thumb, about 10 rats per group is a usually sufficient, although it is preferable to calculate sample sizes using statistical power calculations.

l

Construction of cranial electrodes (Fig. 4) – Stainless-steel screws (00–90  1/16 inch). – Formvar-coated Nichrome wire (0.1000 diameter). – Soldering gun, flux, solder. – Amphinol gold pins.

l

Implantation of cranial recording electrodes – A protocol approved by the institutional committee responsible for oversight of ethical and humane animal care and use. – Stereotaxic device. – Anesthesia: our current protocol involves administration of injectable 5 mg/kg Rimadyl at a 1 mL/kg volume (sc) before beginning surgery, followed by Isoflurane inhalation anesthesia. – Electric shaver. – Ophthalmic ointment. – Betadine surgical scrub. – Surgical tools (scalpel, retractors, wound staples, etc.). – Small drill matched to size of screws. – Screw driver. – Amphenol 9-hole connector with associated gold pins for electrode connectors (Fig. 4). – Dental acrylic (cranioplastic cement).

122

William K. Boyes

– Impedance meter. – Analgesic. l

Creating Visual Stimuli – Computer software capable of generating temporally modulated sine-wave gratings of different spatial frequency and contrast values. Our software package is custom written (22, 23). Other options are available (e.g., http:// psychtoolbox.org/overview.html). – Video monitor (see Note 4). – Photometer.

l

Electrophysiological recording – Faraday cage (grounded). – Shielded cables and connector (paired to headset). – Animal restrainer (head and eyes free for unobstructed view). – Electrophysiological amplifiers and bandpass filters (22). – Computer for experimental control, data collection, signal averaging, and data analysis (23). Other options are available (e.g., https://www.mathworks.com/help/signal/ref/tsa. html). – Ruler. – Black cloth for covering.

3

Methods l

Construction of electrodes (Fig. 4). – Cut nichrome wire to desired length and scrape off the insulation coating from ~2 mm of the both ends. Crimp a gold pin to one scrapped end. – Bend the other scraped end of the wire at a 90 angle and solder this end to the top of a stainless-steel screw, leaving the screw slot opened. – Check continuity from screw to pin with an ohmmeter.

l

Surgical procedures (see Notes 5–10) – Establish survival surgical procedures and obtain approval from the institutional committee responsible for oversight of ethical and humane animal care and use in compliance with applicable guidelines for humane animal experimentation. Approved animal care and use procedures have evolved substantially over time. It is recommended that investigators establish surgical procedures in collaboration with their appropriate institutional authorities (e.g., veterinarian and/or institutional animal care and use committee) to assure

Visual Contrast Sensitivity Testing

123

compliance with current practices rather than rely on previously established procedures. – Anesthetize rat in accordance with approved procedures. – Apply ophthalmic ointment to both eyes to avoid drying of the cornea. – Shave hair from top of head and clean the area with Betadine surgical scrub. – Place head into a stereotaxic frame so that position is secured. – Expose the skull, remove periosteal connective tissue, stop any bleeding using sterile gauze pads and slight pressure. Wipe the skull with a saline-soaked gauze pad and dry the skull with a clean gauze pad. – Mark electrode locations. visual cortex: 4 mm lateral to midline and 1 mm anterior to lambda. reference and ground electrodes: 2 mm anterior to bregma and 2 mm right and left of midline, respectively. – Using sterile drill, gently drill holes in skull for electrode placement. Additional holes may be included for anchor screws to better secure headset to skull. Be careful to gently drill the holes, especially at the base of the skull. The drill should not puncture the dura or damage the underlying brain tissue. – Gently screw the electrodes and anchor screws into the drilled holes. Being careful not to extend below the base of the skull and compress brain tissue. – Using hemostats, snap pin ends of electrodes into designated holes of the Amphenol connector (Fig. 4). – Mix up dental cement and use to build a skull cap securing the electrode assembly. Smooth the exterior of the headset while the cement is still soft. – Once the cement has hardened close the wound using sutures or wound clips. – Remove the animal from the stereotaxic device and allow to recover from anesthesia in a warm and secure chamber. Apply post-surgical analgesia as designated in approved laboratory animal protocols. – Check impedance of electrode assembly using an impedance meter. – House animals singly after surgery to prevent cage-mates from damaging electrode assemblies or electrodes. – Allow approximately 1 week after surgery for recovery and wound healing prior to sensory function testing.

124

William K. Boyes

3.1 Visual System Calibration (see Note 4)

The visual monitor used to present stimuli needs to be calibrated so that there is a linear relationship between input signal voltage and screen luminance. We have used just the green gun of a three-gun color video monitor for presenting visual stimuli. This is (1) in order to simplify calibration (only one of the three cathode ray guns needs to be calibrated) and (2) because the green spectral distribution closely matches the color frequency response spectrum of M-cones and rod photoreceptors, the most common photoreceptors in the rat retina. The voltage-luminance response function (referred to as the gamma function) of most phosphors used in traditional video monitors is typically nonlinear. Therefore, a gamma correction needs to be computed based on the input voltage—screen luminance function established during system calibration. Calibration involves using the computer to generate a steady (unmodulated) spatial square-wave pattern on the screen. The spatial frequency of the stimulus should be low enough (bars wide enough) so that the measurement head of the photometer measures only a single light or dark bar of a square-wave grating stimulus pattern. Alternatively, a non-patterned uniform luminance screen could be used. The calibration involves measuring the luminance at several steps of input signal voltage. Our calibration system interfaces the stimulus computer with the photometer to automatically record the luminance value for each stimulus voltage level presented. Alternatively, the values can be recorded for separate entry. After collecting stimulus voltage-screen luminance values over the range of stimulus levels to be used, a function is fit to the luminance voltage function. Then a mathematical model is generated to linearize the input/ output function (23). The resulting adjusted values should provide a linear relationship between input voltage and output screen luminance. The brightness and contrast adjustments of the monitor can be adjusted to achieve an optimal range of stimulation. It typically is preferred to set the contrast adjustment high and then adjust the brightness to the optimal value for mean screen luminance. Determining a linear range of the gamma-corrected input/output function defines the range of useable luminance and contrast values.

3.2

l

Visual Stimuli

3.2.1 Spatial Properties

Spatial pattern. There are numerous possible choices. We use a vertical sine-wave grating, which is a pattern of vertical bars with a sinusoidal spatial luminance profile. A spatial sinusoid means that there is a gradual transition between the dark and light parts of the screen which would reflect a sine wave if the head of a point photometer measuring luminance were passed along the screen. This is opposed to a square wave (dark and light bars with sharp edges) which would show the sharp transitions from dark to light and back with a moving photometer head. Square waves contain multiple Fourier components and, due to the centersurround receptive fields of visual neurons and their consequent

Visual Contrast Sensitivity Testing

125

spatial frequency tuning, stimulate multiple spatial frequency channels. Spatial sine waves, therefore, are considered more selective and interpretable.

3.2.2 Temporal Properties

3.2.3 Additional Stimulus Parameters

l

Spatial frequency. Select a spatial frequency (bar width) appropriate for the test species or a range of spatial frequency values. The rat has a peak sensitivity to contrast of about 0.16 cycle/ degree visual angle (cpd), and therefore we have used this spatial frequency for most testing applications.

l

Select contrast. In order to construct VEP log-contrast amplitude functions, it is important to test multiple levels of stimulus contrast. The range of contrast levels selected will be limited at the upper end by the highest values of the linear range of the input voltage/output luminance calibration curve. At the lower end, VEPs recorded to low-contrast values (near the perceptual threshold) have low amplitudes and low signal-to-noise ratios. Because of the eventual plotting of VEP amplitudes on linear (amplitude) vs log (contrast) axes, it is beneficial to have log spacing on the contrast values selected.

l

Mode of temporal modulation. There are several possible modes of temporal modulation. We typically prefer sine-wave temporal modulation (the stimulus transitions are gradual) over squarewave temporal modulation (the stimulus transitions are abrupt) because of the same reason to use sine instead of square-wave spatial profiles. The square-wave stimuli contain higher temporal frequency Fourier components that may complicate interpretation of the results.

l

The rate of temporal modulation should be high enough to elicit a semi-sinusoidal response profile. For the rat, this rate is above about 4–5 Hz. This rate yields a strong frequency-double (F2) response amplitude. The temporal modulation rate should not be a direct multiple/divisor of either the line frequency (60 Hz) or the video monitor refresh rate in order to avoid inadvertently signal averaging artifactual electrical noise into the VEP waveform.

l

Luminance level. The overall mean luminance of the pattern should not change during on-off or pattern-reversal modulation. This ensures that the responses reflect changing pattern and not changing luminance (although evoked potentials to changing luminance can also be recorded if that is the goal of the study). We selected a value of about 50 lux for overall mean luminance.

l

Pattern adaptation. Visual neurons adapt rapidly to repeated or continuous visual stimuli. This leads to a rapid reduction in the

126

William K. Boyes

amplitude of VEPs with repeated pattern stimulation. Electrophysiological data are averaged over multiple trials to improve signal-to-noise ratios, yielding an averaged evoked potential. It is recommended to have the subject view mean luminance non-patterned stimuli for adaptation intermixed with stimulus sessions. A general rule of thumb is to have equal amounts of time devoted to adaptation as stimulation. There is a tradeoff between limiting stimulation time to avoid adaptation and increasing stimulation time to provide more trials for signal averaging and improving signal-to-noise levels in the recordings (signal averaging is discussed below). 3.2.4 Electrophysiological Procedures

Signal epoch length. Evoked potentials from steady-state visual modulation take on a sinusoidal form that is readily analyzed with FFT routines. The frequency resolution of an FFT analysis is inversely proportional to the length of the recording epoch. We find that recording 5 s epochs, which gives frequency resolution of 0.2 Hz, gives sharply defined and unambiguous F2 response components. We typically publish only a 1–2 s portion of the 5 s waveform epoch for clarity. Signal averaging. The brain has a considerable amount of spontaneous and ongoing electrical activity—generally referred to as the electroencephalogram (EEG). One approach to separate sensory-related activity from the background ongoing EEG is signal averaging. This involves averaging the electrophysiological data in synchrony with the repeated temporal changes (sometimes accomplished via a “trigger pulse”) so that random ongoing activity is averaged out and activity that regularly accompanies the stimulus is averaged in. The signal-to-noise ratio is increased as the square root of the number of trails averaged. Others have used Fourier analysis of steady-state responses without signal averaging (3). We, however, have had the best results combining signal averaging with FFT analysis of the averaged waveform. We have typically averaged 20 individual sweeps for a steady-state VEP waveform. The decision on how many sweeps to include is a tradeoff between increased signal-to-noise ratios with higher numbers of sweeps averaged, against increased recording time, increased pattern adaption over time, and limiting the number of different stimulus contrast levels (or other parameter of interest) that can be included within any recording session. l

Noise level. Evoked potentials recorded to zero percent contrast (i.e., a non-modulated, non-patterned, mean luminance screen) are thought to reflect the ongoing electrical activity of the brain independent of visual stimuli, and are considered as “noise.” Noise levels can be established by averaging the same number of sweeps and analyzing the data with the same parameters and procedures as for the VEP recording. A noise level value is

Visual Contrast Sensitivity Testing

127

important in understanding the signal-to-noise level, especially of low-contrast responses. VEP recording (see Notes 11–16) l

VEP recording parameters – Filter Bandpass (Hz): 0.1–300 – Sample epoch: 5 s – Analog/digital conversion rate: 1 kHz – Trials averaged: 20

3.3

Data Analysis

l

Placement relative to screen. A restrained rat (see Notes) is placed inside the Faraday cage with its eyes 15 cm from the screen. The optics of the rat eye provide a great depth of field of focus such that everything from 15 cm to infinity can be in focus at the same time. The video monitor is placed outside the Faraday cage to minimize electrical interference in the recordings.

l

Connection. The headset is connected to the shielded cable leading to the electrophysiological amplifiers. The oscilloscope display of the ongoing EEG is checked for connectivity and signal quality.

l

The Faraday cage is covered with a black cloth so that the only light available to the rat comes from the video monitor, which is showing a non-patterned mean luminance adaptation display.

l

Adaptation. The rat is allowed a few minutes in front of the screen in a quiet room to calm and adapt to the screen. Most rats quietly watch the screen.

l

Recording. The stimulus session is started. Several levels of stimulus contrast are presented in a random order.

l

Session monitoring. The operator quietly monitors the rat during the recording session in case eyes are closed or undue amounts of movement artifact are observed. The recording progress is paused, and corrective measures are taken such repositioning the animal or reconnecting the electrode cable as needed.

Spectral Analysis Using FFT. Averaged VEP waveforms can be submitted to readily available FFT routines for spectral analysis. We have used published routines programmed into the signal collection and analysis system (23, 24). The spectral analysis provides frequency components across a range of frequencies, including relatively high frequencies which may not be biologically relevant in these recordings. We typically visualize the frequencies to slightly above 60 Hz in order to assure that the recording is not contaminated by 60 Hz (in the US) line frequency artifact. Spectral

128

William K. Boyes

amplitude is expressed as the square root of spectral power. A prominent spectral peak should be apparent for high-contrast stimuli at twice the rate of temporal modulation (F2), and absent (or nearly so) for the noise level recordings. Higher harmonics (multiples) of F2 may also be apparent in the spectral analysis. Some investigators sum the amplitude at all the harmonic response frequencies. If this approach is selected, then the noise level calculation should also sum noise level recordings at those same frequencies. We have not found an advantage to analyzing the summed harmonics rather than simply using the F2 amplitude. Contrast Amplitude Functions. The VEP F2 amplitude is plotted on a linear scale as a function of log stimulus contrast. Prior to fitting the VEP amplitude log-contrast linear function, it is important to eliminate VEP recordings that are indistinguishable from or below the recording noise level. We eliminate VEP data that are less than the recording noise level plus three times the noise level standard error prior to fitting F2 amplitude log-contrast functions (17). After elimination of noise/level recordings, a linear regression equation is fit to the amplitude log-contrast function. Contrast Thresholds. Contrast threshold may be defined as the zero-amplitude intercept of the contrast amplitude function. Alternatively, threshold may be defined as contrast value where the log-contrast amplitude function crosses the VEP recording noise level. Low-contrast noise level amplitude values can distort slope of regression lines fit to log-contrast amplitude data, driving the intercept of the function to the left on the contrast axis. Because contrast is plotted on a log scale, slight deviations in the slope of the function can have a large influence on the derived intercept value and provide artifactual low measures of contrast thresholds. Contrast Sensitivity. Contrast sensitivity (CS) is defined as the inverse of contrast threshold (CT). CS ¼ 1=CT: Contrast sensitivity functions are typically plotted as the contrast sensitivity value derived for each spatial frequency, over a range of spatial frequencies. Spatial frequency is often plotted on a log scale. Contrast sensitivity functions expressed in this way are typically inverse U-shaped functions with the contrast sensitivity being highest to intermediate spatial frequencies, and lower contrast sensitivity to both lower and higher spatial frequency stimuli (12). Contrast Gain. The slope of the linear function fit to the F2 amplitude/log-contrast function has been interpreted as the “contrast gain.” We fit the log-contrast amplitude function using a linear regression: Y ¼ β0 þ β1 ðXÞ In which Y is the F2 amplitude, X is log contrast, β0 is the intercept, and β1 is the slope parameter.

Visual Contrast Sensitivity Testing

4

129

Notes General Notes: 1. The terms “evoked potential” and “evoked response” are synonymous. 2. Rat strain. Albino strains of rodents are not advised for this procedure. They have poor spatial vision and yield low-quality pattern-elicited VEPs (11). 3. We are describing here the electrode materials and configurations used in our laboratory for over two decades. However, recent descriptions are available for alternative electrodes (25). We have not had experience with these devices, but they may provide better performance than the procedures described here. 4. The present description is based on the tube monitor screens used in our laboratory. Monitors based on other technologies may have different operating characteristics. Problems from surgery: 5. Lost headsets: The headsets may become dislodged for a variety of reasons. (a) Poor adhesion of the acrylic due to not completely cleaning and drying the skull prior to application. (b) Infection: insufficiently sterile procedures. (c) Torque: too much lateral pressure, especially when plugging or unplugging cables for recording. (d) Insufficient anchoring (with screws). (e) Time: the more time after surgery, the more likely headset loss becomes. 6. Broken electrodes, lost connectivity (a) Loose crimping of pin to wire. (b) Poor soldering of wire to screw. (c) Too much pressure on electrode during surgery, especially when inserting pin to headset connector. 7. Electrodes too deep—focal lesions (a) Electrode should not extend below the bottom of the skull as this can damage the cortex. 8. Infection (a) Review surgical procedures for cleanliness. (b) Keep area surrounding surgical procedures clean and draped. (c) Sterilize instruments between procedures.

130

William K. Boyes

9. Damage from group housing (a) Rats that are group housed will chew on cage-mate’s headsets and destroy the electrodes. It is imperative to singlehouse rats after cranial implantation. 10. Loss of body weight/ill health (a) Rats should be followed post-surgery for general health and body weight. It is normal to lose a few grams body weight after surgery. Excessive weight loss is a sign of infection or other poor consequences of surgery. Animals that do not thrive after surgery should be eliminated from the study. Problems during recordings: 11. Eyes closed. Sometimes rats close their eyes during recordings. This is especially true if they are stressed. It is possible to wait for them to calm down. It may be necessary to adjust their restraints to improve comfort. It is helpful to be calm and confident when handling the rats to lower their initial stress response to the restraint. 12. Unplugged/lost headset. Monitor the EEG on an oscilloscope during all recording sessions. The signal from intact electrodes and a connected animal will be a familiar EEG pattern. Noise in the recording is immediately apparent as large 60 Hz patterns due to induction and amplification of the building electrical activity. 13. Escape restraint. Rats are motivated escape restraint. We have found that plastic de-capitation cones work well as disposable single-use restrainers. The narrow end of the plastic cone is cut open to expose the head, (including eyes and ears), prior to placing the rat into the cone. The rat’s toes can be covered with small pieces of tape to prevent clawing their way out of the device. The rat is placed into the cone and the area around the neck is taped to be snug, but not to impair breathing. A rectal temperature probe is inserted and taped in place. The rear end of the cone is taped shut with the rat’s rear legs facing backwards. The rat in the cone is placed on a specially designed holder made from a half-longitudinal section of PVC plumbing pipe, mounted on an adjustable stand. 14. Excessive movement/artifact rejections. Some rats display movements, such a struggling against the restraint or chewing during recording sessions which cause large electrical artifacts in the EEG recordings. The data acquisition program should include artifact rejection capability such that signals which exceed the digital-analog converter input window, or a designated percentage of it, are rejected prior to inclusion in the average evoked potential. Calming the rat, reconfiguring the

Visual Contrast Sensitivity Testing

131

restraint, or waiting can reduce movements and their associated artifacts. 15. 60 cycle noise. As for any neurophysiological recording, care must be taken to eliminate 60 Hz noise. The methods to do this are well-described elsewhere, and involve use of a grounded Faraday cage, proper grounding of equipment, avoiding ground loops and shielding of electrode cables. The use of 60 Hz notch filters is discouraged as they can distort the signal and are a poor substitute for proper grounding and elimination of other sources of artifacts. 16. Chromodacryorrhea. Some rats exhibit a red porphyrin containing lacrimal discharge around their eyes referred to as chromodacryorrhea. This can indicate a stress response. Because it can interfere with vision if covering the eye, chromodacryorrhea should be gently removed with a cotton swab while recordings are paused. Measures to reduce stress are advisable.

Acknowledgments The author thanks David Herr, Jordi Llorens, and an anonymous reviewer for helpful comments on an earlier version of the manuscript. The author thanks Mary Gilbert for initiating the experiments depicted in Fig. 8. The author also thanks Garyn Jung for assistance and Chuck Gaul for photography in creating Fig. 4. The surgical and electrophysiological procedures were based on those originally developed by Robert S. Dyer with technical innovations by Mark Bercegeay. This document has been subjected to review by the National Health and Environmental Effects Research Laboratory and approved for publication. Approval does not signify that the contents reflect the views of the Agency, nor does mention of trade names or commercial products constitute endorsement or recommendation for use. References 1. Herr DW, Boyes WK (1995) Chapter 9— Electrophysiological analysis of complex brain systems: sensory-evoked potentials and their generators. In: Chang LW, Slikker W (eds) Neurotoxicology. Academic Press, San Diego, pp 205–221. https://doi.org/10.1016/ B978-012168055-8/50013-3 2. Norcia AM, Appelbaum LG, Ales JM, Cottereau BR, Rossion B (2015) The steady-state visual evoked potential in vision research: a review. J Vis 15:4

3. Regan DEDE (1989) Human brain electrophysiology : evoked potentials and evoked magnetic fields in science and medicine. Elsevier, New York 4. Kothari R, Bokariya P, Singh S, Singh R, Comprehensive A (2016) Review on methodologies employed for visual evoked potentials. Scientifica (Cairo) 2016:9852194 5. Creel DJ (2019) Visually evoked potentials. Handb Clin Neurol 160:501–522

132

William K. Boyes

6. Tseng HC et al (2015) Visual impairment in an optineurin mouse model of primary openangle glaucoma. Neurobiol Aging 36:2201–2212 7. Demyanenko GP et al (2011) NrCAM deletion causes topographic mistargeting of thalamocortical axons to the visual cortex and disrupts visual acuity. J Neurosci 31:1545–1558 8. Strain GM, Tedford BL, Gill MS (2006) Brainstem auditory evoked potentials and flash visual evoked potentials in Vietnamese miniature pot-bellied pigs. Res Vet Sci 80:91–95 9. Mitzdorf U (1987) Properties of the evoked potential generators: current source-density analysis of visually evoked potentials in the cat cortex. Int J Neurosci 33:33–59 10. Schroeder CE, Tenke CE, Givre SJ, Arezzo JC, Vaughan HG Jr (1991) Striate cortical contribution to the surface-recorded pattern-reversal VEP in the alert monkey. Vis Res 31:1143–1157 11. Boyes WK, Dyer RS (1983) Pattern reversal visual evoked potentials in awake rats. Brain Res Bull 10:817–823 12. Boyes WK (1994) Rat and human sensory evoked potentials and the predictability of human neurotoxicity from rat data. Neurotoxicology 15:569–578 13. Birch D, Jacobs GH (1979) Spatial contrast sensitivity in albino and pigmented rats. Vis Res 19:933–937 14. Silveira LC, Heywood CA, Cowey A (1987) Contrast sensitivity and visual acuity of the pigmented rat determined electrophysiologically. Vis Res 27:1719–1731 15. Bobak P, Bodis-Wollner I, Marx MS (1988) Cortical contrast gain control in human spatial vision. J Physiol 405:421–437 16. Bonds AB (1991) Temporal dynamics of contrast gain in single cells of the cat striate cortex. Vis Neurosci 6:239–255

17. Boyes WK, Degn L, George BJ, Gilbert ME (2018) Moderate perinatal thyroid hormone insufficiency alters visual system function in adult rats. Neurotoxicology 67:73–83 18. Boyes WK et al (2003) Dose-based duration adjustments for the effects of inhaled trichloroethylene on rat visual function. Toxicol Sci 76:121–130 19. Boyes WK et al (2014) Neurophysiological assessment of auditory, peripheral nerve, somatosensory, and visual system functions after developmental exposure to ethanol vapors. Neurotoxicol Teratol 43:1–10 20. Boyes WK et al (2005) Momentary brain concentration of trichloroethylene predicts the effects on rat visual function. Toxicol Sci 87:187–196 21. Boyes WK et al (2016) Toluene inhalation exposure for 13 weeks causes persistent changes in electroretinograms of long-Evans rats. Neurotoxicology 53:257–270 22. Herr DW et al (2016) Neurophysiological assessment of auditory, peripheral nerve, somatosensory, and visual system function after developmental exposure to gasoline, E15, and E85 vapors. Neurotoxicol Teratol 54:78–88 23. Hamm CW, Ali JS, Herr DW (2000) A system for simultaneous multiple subject, multiple stimulus modality, and multiple channel collection and analysis of sensory evoked potentials. J Neurosci Methods 102:95–108 24. Bergland GD, Dolan MT (1979) Fast Fourier transform algorithms. In: Weinstein CJ (ed) Programs for digital signal processing. John Wiley & Sons, Inc, New York, pp 1.21–1.2-18 25. Tian L et al (2019) Large-area MRI-compatible epidermal electronic interfaces for prosthetic control and cognitive monitoring. Nat Biomed Eng 3:194–205

Chapter 7 Electrophysiological Assessments in Peripheral Nerves and Spinal Cord in Rodent Models of Chemotherapy-Induced Painful Peripheral Neuropathy Susanna Park, Cynthia L. Renn, Justin G. Lees, Susan G. Dorsey, Guido Cavaletti, and Valentina A. Carozzi Abstract Chemotherapy-induced peripheral neuropathy (CIPN) is a severe side effect related to anticancer treatment, typically characterized by sensory symptoms including numbness, tingling in the distal extremities and neurophysiological impairments. CIPN is often painful, which is identified by adding a second “P” to the acronym. The incidence of CIPN is variable depending on the drug, pre-existing neuropathy, and clinical history, but generally increases with the cumulative dose and can also persist after treatment discontinuation. In the last 30 years, many rodent models of CIPN have been developed reproducing the clinical features of the pathology, useful to study the mechanisms of pathogenesis and test neuroprotective strategies. In this chapter, we will focus our attention on sensitive and reproducible methods to study the pathophysiology of chemotherapy-induced painful peripheral neuropathy (CIPPN), in animal models. In particular, we describe the techniques to record nerve conduction velocity and nerve excitability parameters in peripheral nerves and the electrical activity of wide dynamic range neurons of the dorsal horn of the spinal cord in mice, as parameters of evaluation of nerve function and painful neuronal sensitization, respectively. Our intent is to provide the reader with guidelines on how to prepare and manage the animals according to the 3Rs (Reduction, Refinement, and Replacement) principles, how to record neuronal activity and analyze resulting data and describe common technical problems and appropriate solutions. These protocols can also be useful to study peripheral nerve damage and pain of other origins, such as traumatic injury, inherited, or acquired neuropathies. Key words Chemotherapy, Neuropathic pain, Peripheral neuropathy, Electrophysiology, Peripheral nerves, Wide dynamic range neurons, Spinal cord

Susanna Park and Cynthia L. Renn equally contributed to this work. Jordi Llorens and Marta Barenys (eds.), Experimental Neurotoxicology Methods, Neuromethods, vol. 172, https://doi.org/10.1007/978-1-0716-1637-6_7, © Springer Science+Business Media, LLC, part of Springer Nature 2021

133

134

1

Susanna Park et al.

Introduction

1.1 Clinical Features of ChemotherapyInduced Painful Peripheral Neuropathy (CIPPN)

Chemotherapy-induced peripheral neuropathy (CIPN) is a prominent side effect of cancer treatment, affecting patients treated with commonly used chemotherapies. Chemotherapy classes including taxanes, platinum-based agents, vinca alkaloids, thalidomide, and bortezomib analogs are associated with the development of CIPN. CIPN typically produces sensory symptoms, most prominently including numbness and tingling in the distal extremities of the hands and feet and can include pain (CIPPN) [1]. In severe cases, these symptoms produce functional disability, leading to difficulty with walking, balance, fine motor skills, and ultimately activities of daily life [2]. Severe CIPN often leads to dose reduction or premature cessation of treatment, which may affect long-term outcomes. Further, CIPN can produce long-lasting symptoms, leading to reduced quality of life in cancer survivors [3]. There remains no neuroprotective therapy to reverse peripheral nerve damage due to CIPN. The presentation of CIPN can vary between chemotherapies and can include motor or autonomic nerve involvement in addition to sensory involvement [1]. Motor nerve involvement can produce weakness while autonomic nerve involvement can include gut dysfunction or orthostatic hypotension. The most prominent electrophysiological finding on nerve conduction studies (NCS) is the reduction or loss of sensory compound action potentials [4], highlighting an axonal, sensory predominant neuropathy with most chemotherapy types. While the most common reported sensory symptom of CIPN is tingling and numbness, depending on the agent, neuropathic pain can also occur in 25–40% of patients [5, 6]. Neuropathic pain in CIPPN, characterized by burning sensations, may have an additional adverse effect on quality of life. Typically, the incidence and severity of CIPPN increases with increasing cumulative dose of neurotoxic chemotherapy [2]. However, some chemotherapies are also associated with acute neurotoxic syndromes—most notably oxaliplatin, which produces acute cold-triggered tingling, dysesthesia, and cramps immediately following infusion [7]. However, there is significant variability in clinical expression and severity between individuals, suggesting that patient-specific or pharmacogenetic risk factors are also important [8].

Electromyographically Assessments of Chemotherapy-Induced Painful. . .

1.2 Animal Models and Methods of ChemotherapyInduced Painful Peripheral Neuropathy (CIPPN) Investigation 1.2.1 Animal Models of CIPPN

135

Rodent models have been the most commonly utilized experimental models for the study of CIPPN to date. Many preclinical rodent (both rat and mouse) models [9–18] have been established during the last 30 years faithfully mimicking the clinical features of CIPPN. Because the majority of neurotoxic chemotherapy drugs do not cross the blood–brain barrier, most preclinical studies have been focused on the study of the function and structure of peripheral nerves and dorsal root ganglia (DRG). Direct toxic effects of chemotherapy drugs have been observed in peripheral axons, in primary afferent sensory neurons, and in peripheral support cells such as satellite cells in the DRG and the Schwann cells in peripheral nerves. Preclinical animal studies showed that chemotherapyinduced cellular abnormalities in both the nerve and glial cells lead to structural damage, to alteration in neuronal-glial cross talk, and finally to the loss of peripheral nerve function. The specific mechanisms producing nerve degeneration vary in relation to the different mechanism of action of each chemotherapy drug. Moreover, animal studies have shown that chemotherapy-induced neurotoxic damage in the peripheral sensory fibers (A-Beta, A-Delta, and C fibers) is not peripherally confined but is able to induce indirect alterations of transmission through the somatosensory system and spinal cord [9, 10, 12, 13]. This generates an excess of spinal neuron excitability, which coupled with a decrease of inhibition, results in spinal hyper-excitability, increased mechanical and/or thermal sensitization typical of neuropathic pain conditions. As a general consideration, acute models of CIPPN (those in which the chemotherapy drugs are injected only once or few consecutive times) faithfully reproduce the early and transient painful symptoms after chemotherapy treatment (e.g., the cold hyperalgesia reported by patients few hours after oxaliplatin injection). Chronic models of CIPPN (those in which the chemotherapy drugs are repeatedly injected for several weeks) are characterized by established peripheral nerve lesions that underlie the typical features of painful peripheral neuropathy (neurophysiological abnormalities, decrease in intra-epidermal nerve fiber density, neuropathological alterations into peripheral nerves, DRG, spinal roots and mechanical allodynia, hyperalgesia or hypoesthesia, thermal hyper- or hypoalgesia). These models are characterized by ectopic discharges and spontaneous action-potential firing in primary sensory afferents, which accordingly produce mechanical and/or thermal sensitization in the central nervous system. In recent decades, many mouse and rat models of acute and chronic CIPPN have been developed and phenotypically characterized [9–28].

136

Susanna Park et al.

1.2.2 Methods of Experimental CIPPN Investigation

In addition to the establishment of reliable rodent models of CIPPN, the setup of sensitive and reproducible methods of investigation of neurotoxicity and pain are mandatory for the success of an experimental paradigm.

Morphological Methods

Light and electron microscopy analysis are useful tools to investigate the presence and the subcellular origin of the structural damage induced by chemotherapy on its peripheral target sites (DRG, peripheral nerve fibers). Briefly, most chemotherapy drugs produce axonal degeneration with occasional additional damage to myelin, as well as alterations in the structure of DRG sensory neurons with organelle vacuolization and cytoplasmic dark inclusions. In addition, satellite cells can be damaged and increased in number. Sensory neurons can also undergo atrophy or hypertrophy depending on the chemotherapy drug used. Similarly, the density of peripheral nerve fibers as well as the intra-epidermal nerve fibers density can be reduced [10, 14, 15]. Apoptotic and/or degenerative processes induced by chemotherapy can be identified via staining (i.e., TUNEL, FluoroJade) as well as immune-histochemical labeling against specific markers (together with molecular biology and biochemical assays) to help in the identification of the cellular pathways leading to neurotoxicity [12, 15].

Electrophysiological Methods: Nerve Electrophysiology

Nerve electrophysiological methods are minimally invasive techniques which enable assessment of peripheral nerve function in situ. Nerve conduction studies (NCS) are widely utilized in the clinical neurology setting as the gold standard technique to assess peripheral neuropathy [29]. NCS assess the properties of the fastest and largest conducting axons but are insensitive to small, unmyelinated axonal dysfunction. The major components of NCS include assessment of the size of the compound action potential (amplitude) and assessment of the speed of conduction (nerve conduction velocity NCV). These parameters can be collected from different peripheral nerves, stimulating directly through the skin or using needle electrodes. Compound muscle action potentials (CMAPs) can be recorded from innervated muscles following electrical stimulation of a motor nerve. Similarly, compound sensory action potentials (CSAPs) can be recorded following electrical stimulation of sensory fibers from a site within the distribution of the nerve [30]. NCS can reveal evidence of axonal damage or demyelination [31]. Broadly, reduction in compound action potential amplitudes can demonstrate the development of axonal neuropathy and changes in conduction velocity can indicate demyelination. Conduction slowing can be demonstrated via slowed conduction velocity, prolonged latency, or increased temporal dispersion. Axonal damage can be characterized by decreased distal amplitudes without prominent conduction slowing although mild conduction velocity slowing can occur in moderate to severe polyneuropathy [32]. Importantly,

Electromyographically Assessments of Chemotherapy-Induced Painful. . .

137

very similar or analogous techniques can also be utilized in animal models of CIPPN, which provides an important way to compare disease models to the clinical setting. While NCS examine the amplitude and conduction velocity of the largest conducting fibers, nerve excitability studies utilize patterns of stimulation to examine modulation of excitability in response to impulse conduction. These patterns have been linked to ion channel function and membrane potential [33, 34]. Nerve excitability studies have also been established in animal models, examining both sensory and motor nerve excitability in rats and mice [35–37] and accordingly can be compared between animal models and the clinical setting. Neurophysiological studies can be undertaken in sensory, motor, or mixed nerves in rodent models. Motor nerve conduction studies are often undertaken in the sciatic nerve, with the stimulating electrodes at the sciatic notch and the recording from the foot or leg [38]. To generate conduction velocity, a second, more proximal stimulation site is selected. Sensory nerve conduction studies are often undertaken in the tail in rodent models, stimulating at the base of the tail and recording more distally along the caudal nerve [39]. However, there are many nerves and recording sites which have been utilized for neurophysiological recordings across different animal models of peripheral neuropathy. In the animal model setting, it is important to note that there is some overlap between these phenomena and the time required for the development of axonal loss means that this is not always demonstrated in models. It is often more difficult to reliably measure action potential amplitude compared to NCV, so it can be difficult to dissociate between a primary effect on myelin compared to axonal damage [40]. Electrophysiological Methods: Spinal Cord Electrophysiology

As mentioned previously, an increased activity in the peripheral sensory neurons leads to an increased activity in the spinal dorsal horn neurons, which then translates to an increased neuronal activity in the higher centers of the CNS [41]. There are two types of neurons in the spinal dorsal horn that transmit nociceptive information. One type is represented by the nociceptive-specific neurons that respond only to noxious stimuli in the tissue-damaging range of intensity and will remain silent in response to innocuous stimuli [42]. The second type is represented by the wide dynamic range (WDR) neurons that are located in the deep layers of the spinal dorsal horn, predominantly lamina V [43, 44]. The WDR neurons were first identified by Mendell in 1966 [45] and later were well characterized in the 1980s [46–48]. Unlike the nociceptive-specific neurons that have a very high threshold of activation, the WDR neurons are responsive to innocuous low-intensity stimuli from A-Beta fibers as well as to noxious high-intensity stimuli from A-Delta fibers [43]. WDR neurons are also activated by stimuli

138

Susanna Park et al.

from intrinsic interneurons that transmit signals from C-fiber terminals in laminae I and II [43]. In a persistent pain state, the ongoing nociceptive stimuli from primary afferent neurons lead to the development of wind-up, which induces hyperexcitability of the WDR neurons, neuroplasticity, and eventually central sensitization [44]. When WDR neurons become hyperexcitable, they produce an exaggerated response to innocuous stimuli from A-Beta fibers. These stimuli are then perceived as painful rather than innocuous, which is the basis for the phenomenon of allodynia that is associated with neuropathic pain [43, 44]. WDR neurons represent a good target for electrophysiological examination of whether central changes have occurred in response to persistent pain, as in CIPPN [49, 50]. In the absence of central neuronal plasticity, WDR neurons generate action potentials in a graded response to varying stimulus intensity [43]. Thus, if changes in the function of primary afferent fibers are detected, but the stimulus-response of WDR neurons remains within the normal range, then it is not likely that central neuronal plasticity or central sensitization have developed [43]. However, if the WDR evoked response and after discharge to innocuous stimuli is significantly increased, then that is evidence that neuronal plasticity and possible central sensitization have developed [51]. Electrophysiological recording of WDR neurons is an excellent tool for studying the neurophysiological development of CIPPN since changes in the response properties of WDR neurons persist after a painful stimulus similar to the persistence of the psychophysical ratings of pain intensity and unpleasantness [42]. Also, given that the response properties of WDR neurons are similar across a variety of mammalian species [42], recording the response of WDR neurons in the development of CIPPN in rodents can be an indication for what may occur in humans.

2

Protocols Described There are many experimental approaches to investigate CIPPN. This chapter describes the protocols for the recording of NCV and nerve excitability parameters in peripheral nerves and the electrical activity of WDR neurons of the spinal cord of mice. Other experimental approaches are beyond the scope of this chapter but are addressed in detail by Bruna and collaborators [52] and Hoke and collaborators [38]. All the procedures described in this chapter comply with local governmental regulation concerning the care of laboratory animals, in vivo research protocols and the 3Rs (Reduction, Refinement, and Replacement) principles. Although different instrumentation is described in the two protocols, some of the equipment can be employed for both

Electromyographically Assessments of Chemotherapy-Induced Painful. . .

139

Fig. 1 (a) Equipment and materials for nerve excitability experiments. Panels (b–d) depict example traces of nerve excitability recordings from the mouse caudal nerve, illustrating (b) threshold electrotonus, (c) recovery cycle and (d) current-threshold relationship waveforms. Excitability traces provided by PG Makker

protocols (including the anesthesia system, amplifier, noise eliminator). Similarly, supplies including alligator clips, connection cables, aluminum foil, syringes, hemostatic sponges, cotton swabs, and animal shavers can be useful for both procedures.

2.1 Materials and Methods for Electrophysiology in Peripheral Nerves

In this section, we describe the procedures to record NCV and nerve excitability parameters from mice [25, 37]. The instrumentation required is illustrated in Fig. 1.

2.1.1 Materials

Preparation of the animal l

Isoflurane 2% anesthetic system with vaporizer and oxygen supply equipped with an anesthetizing box and nose-cone mask fixed to heating block.

l

Homeothermic blanket system adjusted by rectal probe to maintain core body temperature at 37
C. Electrophysiological recordings

l

QtracS software (Copyright Institute of Neurology, UCL, available from Digitimer Ltd.) for data acquisition.

l

Isolated amplifier—D440-2; Digitimer Ltd. (Welwyn Garden City, UK).

l

Digitizer—16-bit data acquisition National instruments (Austin, Texas).

l

Isolated stimulator Model 2200 Analog Stimulus Isolator; A-M systems (Carlsborg, Washington).

system—USB-6251;

140

Susanna Park et al.

2.1.2 Methods

l

Noise eliminator 50/60 Hz, eliminate 50/60 Hz noise and harmonics without filtering (KF Technologies).

l

Disposable non-polarizable Ag/AgCl ring electrodes (The Electrode Store; Buckley, Washington).

l

Platinum needle electrodes (Natus; San Carlos, California).

Preparing the mouse 1. Mouse (see Note 1) is removed from the holding cage and placed into anesthetizing box with intake of 2% isoflurane in oxygen. Wait until mouse is completely unconscious (~4 min). 2. Remove mouse from anesthetizing box and transfer to nosecone apparatus supplied with intake of 2% isoflurane in oxygen. 3. Used toe-pinch test to ensure that mouse is completely unconscious. 4. Ensure mouse is in a supine position on top of a heating mat covered with paper toweling. Heating mat is temperature adjusted by rectal probe which is inserted into mouse rectum for duration of procedure maintaining core body temperature at 37
C. 5. Immobilize mouse by placing a folded paper towel over upper body torso and secure to underlying mat with tape. Electrophysiological study setup 1. To stimulate the caudal nerve, Ag/AgCl ring electrodes (see Note 2) connected to the stimulator are attached. The anode electrode is placed on the left rear limb around the ankle and the cathode electrode is placed around the base of the tail. 2. To record from the caudal nerve, platinum needle recording electrodes connected to amplifier are fitted to record compound muscle action potentials (CMAPs) and sensory nerve action potentials (SNAPs): (a) For CMAP recording the active electrode is inserted into the tail muscle 20–30 mm distal to the cathode. The reference electrode is inserted into the tail 8 mm distal to the active electrode. (b) For antidromic SNAP recording the active electrode is inserted into the tail-skin 65–70 mm distal to the cathode. The reference electrode is inserted into the tail-skin 8 mm distal to the active electrode. 3. For both CMAP and SNAP recording a grounded ring electrode is placed around the tail distal to the cathode and proximal to the active recording electrode.

Electromyographically Assessments of Chemotherapy-Induced Painful. . .

141

Recording of action potential amplitude and conduction velocity (NCV) 1. The optimal stimulation parameters should be determined to generate a maximal response, for example, 10 mA stimulus intensity, 0.75 s frequency of stimulation, and 0.04 ms stimulus duration from Renn et al. 2011 [13]. 2. To record the maximal compound action potential amplitude, increase the stimulus intensity until the amplitude of the compound action potential does not further increase. In order to ensure that supramaximal stimulation has been reached, further increase the stimulus intensity by 20%, ensuring that there is no further increase in the compound action potential amplitude. Record the values as the maximal CMAP or SNAP as appropriate. 3. Caudal nerve NCV can be recorded and calculated as the ratio of the distance between the stimulating and recording electrodes and the latency between the stimulus artifact and the onset of the action potential (Fig. 2). Supramaximal stimulation, as above, should be utilized to determine NCV. The NCV is calculated in m/s.

Fig. 2 Experimental setup for nerve excitability experiments for (a) sensory nerves and (b) motor nerves. Photographs taken by PG Makker

142

Susanna Park et al.

It should be noted that for the calculation of sensory NCV, only one stimulating site is required. However, for determination of motor NCV, two stimulating sites are required, one distal and one proximal. Distal motor latency is subtracted from the proximal motor latency to calculate the nerve conduction time between sites. In conjunction with the distance between the two sites, this can be used to determine NCV for motor nerves. Initiating recording of excitability 1. Excitability studies are undertaken using a constant-current stimulator, with computer software QTRAC (Institute of Neurology) used to control the stimulation, recording and for processing data. 2. Multiple excitability measurements can be recorded with specialized semi-automated protocols, (e.g., the TROND protocol, developed to record multiple excitability measurements via threshold tracking; Kiernan et al. 2019 [33]). 3. Initially, the stimulus current (stimulus duration 1 ms motor, 0.5 ms sensory) is manually increased until a maximal response is obtained. 4. A stimulus-response curve is automatically generated, utilizing the pre-set maximum current level and incrementally decreasing the stimulus current by 2% steps until the response is absent. 5. The target amplitude for threshold tracking is set to 40% of the maximum response for both motor and sensory recordings. 6. For sensory recordings, additional averaging is required, using the small sensory option in QTracS. Nerve excitability protocols 1. Multiple excitability parameters are semi-automatically collected, including: (a) Strength duration time constant: altering stimulus durations (0.2–1 ms for motor; 0.1–0.5 for sensory). (b) Threshold electrotonus: using subthreshold polarizing currents in both hyper- and depolarizing directions (20%, 40%, 70%, 100% of control threshold). (c) Current-threshold relationship: Effect of a 200 msec polarizing current (+50% to 100% of control threshold). (d) Recovery cycle: Changing interstimulus intervals between an initial supramaximal conditioning stimulus and the target potential (from 1.3 to 200 msec; see Note 3).

Electromyographically Assessments of Chemotherapy-Induced Painful. . . 2.1.3 Data Analysis

2.1.4 Notes

143

Nerve excitability study data can be analyzed and visualized with the QTracP program utilizing multiple excitability measure files (MEM). 1. Both male [53] and female [54] rodents have been shown to demonstrate changes in NCV due to chemotherapy treatment. However, few studies have included both sexes to directly compare the sexes using the same chemotherapy dosage regimen. However, the strain of mouse does appear to make a difference in the degree of electrophysiological changes with chemotherapy treatment—with DBA2J mice showing significant alterations following paclitaxel treatment but C57BL/6 mice remaining unaffected [55]. Further, most of these experiments are performed in young adult rodents. The obvious advantage of using more mature mice is that the tail is longer and therefore when recording from caudal nerve there is more distance to separate the stimulation location and the sensory nerve recording location giving better separation from the stimulation artifact [25]. 2. In the majority of studies, needle electrodes are used for both stimulation and recording of CMAP and SNAP [55]. An advancement for stimulating the caudal muscle and nerve is the incorporation of ring electrodes that create an electrical field around the entire section of the tail being stimulated and therefore provide and more even and stronger stimulation for distal recording [25]. 3. In the recovery cycle when the interstimulus intervals are close together (95%) damage to the olfactory epithelium [90]. 3.4 Neurophysiologic and Brain Imaging Approaches to Assess Olfactory Function

The electroolfactogram (EOG) assesses the collective response of groups of olfactory neurons following odorant exposure. The EOG has been primarily used to study spatial and temporal responses and help map the distribution of olfactory receptor genes in zones in the rodent olfactory epithelium [91, 92]. The EOG has been used both in vivo and in vitro using organotypic cell cultures [93, 94]. To date, application of the EOG to rodent toxicity studies has been limited [95, 96]. A useful video demonstration of the EOG is available [93]. The EOG directly measures olfactory neuron responses to an odorant. An alternative approach to assess the integrity of the olfactory system would be to assess odor-induced changes in the electroencephalogram (EEG). In such studies, EEG signals are recorded during odorant presentation [97, 98]. A related method, magnetoencephalography, has also been used to evaluate human olfaction [99]. In humans, surface electrodes are used to collect the EEG signal and studies using a combination of both EEG and functional MRI have been reported [100, 101]. In rodents electrodes implanted in the olfactory bulb are typically used [102, 103]. As is the case with the EOG, odor-induced EEGs have been infrequently used to assess olfaction in rodents following toxicant exposure. Brain imaging has received increased use in the evaluation of smell disorders in people. These methods have been used to identify cortical structures associated with olfactory stimulation and

Assessment of Olfactory Toxicity in Rodents

229

have also been introduced into clinical practice. Both positron emission tomography [104–106] and functional MRI [107–109] have been applied to the study of olfaction in people. Many of these approaches have been used in rodents as well [110–115]. Other more specialized approaches including optogenetics and two-photon imaging have been used to study rodent olfaction [116–118]. As these methods become more widely available, we anticipate that they may be increasingly used in future rodent toxicity studies.

4

Notes Several technical issues need to be considered when assessing olfaction in rodents. Training of staff to detect cues associated with olfaction is critical. Inter-individual variability in assessment of olfaction should be minimized. One approach is to use the same individual to assess olfaction in animals. Blinding of the observer to animal treatment groups can also reduce bias. Species selection often depends on the research objectives of the toxicity study. Regulatory toxicology studies often rely on rats whereas more investigative studies often favor mice due to availability of knockout mice and other gene targeted models. Sex- and strain-differences in rodent olfaction are recognized and these factors may affect results in toxicity studies. For example, aging in mice is associated with decreased numbers of sensory olfactory neurons, reduced regenerative capacity of the olfactory epithelium, and decreased olfactory ability [119, 120]. It is plausible that the aged olfactory system may be at greater risk to toxicant exposure. Female mammals generally have a better sense of smell than males [119–121], and this can be further influenced by the reproductive status of the animal [122, 123]. The method of chemical exposure, inhalation, intranasal instillation, or systemic delivery, depends upon the toxicant of interest and the technical capabilities of the laboratory performing the work. In mice, vanillin is neutral with respect to odor preference— mice show neither attraction nor revulsion to the odor [124]. Short-term (20 min) exposure of mice to vanillin has minimal impacts on their performance in an elevated plus maze, open field, Y-maze, among other behavioral tests. Short-term exposure to vanillin was associated with both decreased hot plate response and reduced forelimb grip strength [124]. The impact of other odorants on neurobehavioral test performance by rodents remains incompletely understood. Concerns might be raised regarding whether or not the use of a limited number of odorants to study rodent olfaction is sufficient given the large number of odors that exist in the environment. Studies performed in humans have shown

230

David C. Dorman and Melanie L. Foster

that as few as three odors, namely cinnamon, banana, and fish odor may be needed to quickly identify an individual as having normal olfactory function [125]. Stimulus control can be a daunting technical challenge since unlike sound or light, odor cues cannot be easily generated or turned on or off. Extraneous odors from personal care products, cleaning supplies, and other materials should be avoided. Likewise, maintaining a constant odorant concentration is difficult since the odor stimulus can fluctuate over time [126]. Odor cues in some suprathreshold tests of olfaction are presented using cotton balls or other media that are impregnated with solutions that contain the odorant of interest. Other tests of rodent olfaction rely on the generation of an atmosphere containing the odorant of interest. These atmospheres are often produced by placing the odorant of interest in water or another liquid vehicle and then relying on vaporization of the odorant to produce the atmosphere. Saturated solutions of the odorant are often used as the odor source resulting in suprathreshold exposures. Air odorant concentration derived from solutions of an odorant depends upon the concentration of the odorant in the vehicle, temperature, and the odorant’s vapor pressure. In other cases, Tedlar bags or other inert materials containing known amounts of the odorant of interest mixed with air or nitrogen can be used to generate the odor source [49]. Air odorant concentrations delivered to the animal or test system can be modified by mixing the odor source air with a clean (no odorant) air source producing subthreshold concentrations useful for the measurement of an olfactory threshold. The odorant generation and delivery system should use nonreactive components that do not contribute to the odor signature (e.g., Teflon). Separate odor generation systems may be needed to reduce cross-contamination when multiple odors are being used. Nominal air concentrations of the odorant can be confirmed using gas chromatography and mass spectrometry or other analytical chemistry methods. Once the odorant containing atmosphere has been generated it needs to be presented to the breathing zone of the animal. Commercially available microprocessor controlled “odor ports” are available that can present the odorant to an animal performing a behavior in an operant chamber. Odorant can also be presented in other ways including infusing a test chamber with the odorant of interest. Residual presence of an odorant in a test apparatus or an animal room may confound results—therefore, some effort to reduce residual odor (e.g., using an exhaust system) may be helpful. Using your sense of smell to determine when an odor is no longer present is unreliable since olfactory thresholds for many odorants are appreciably lower in rodents when compared with people. Likewise, test apparatuses should be cleaned with distilled water or other unscented cleaning solutions to reduce distracting odors in a test system.

Assessment of Olfactory Toxicity in Rodents

231

It can also be challenging to recognize when an animal responds to the odorant. Active sniffing in small rodents with high respiratory rates is difficult to perceive visually. In many cases, tests of rodent olfaction rely on the expression of a second behavior (e.g., uncovering of a buried food pellet, lever pressing) that is presumed to indicate that the animal has detected the odor. Animals that can no longer perform these secondary behaviors may be misclassified as having a decreased olfactory ability. It is therefore critical that controls be incorporated into the study to confirm that the animal’s ability to perform the task is unaffected by toxicant exposure. In addition, some commonly used test methods rely on the animal responding to an odor cue in order to receive a food reward. In these cases, animals are often maintained at 80 to 90% of their normal body weight or held off food overnight prior to testing. Chemicals that decrease motivation may result in spurious results on these assays.

5

Conclusions In closing there are several “take home” messages that we want to impart to the reader. Evaluation of olfactory function in rodents occurs widely in neuroscience; however, approaches used by neuroscientists have rarely been adopted by toxicologists. There is an important opportunity for future application of these methods to neurotoxicology. For this and other reasons, olfaction and toxic effects on the rodent olfactory system remains underappreciated. Most studies used to evaluate the rodent olfactory system rely on histologic evaluation of multiple sections of the nasal cavity and relatively simple tests of olfaction. To date many of these behavioral tests have been unable to demonstrate a lack of olfactory function even in animals with significant pathology in the olfactory system. For example, extensive injury and loss of sensory olfactory neurons as assessed using histopathology is often required before a functional deficit can be observed [49, 90, 127]. Other studies have shown that near total ablation of the rat olfactory bulb did not impair an animal’s ability to detect and discriminate between odors [128]. These findings suggest that a tremendous “reserve capacity” in the rodent for the sense of smell likely exists. In humans, we rarely have information regarding the structural integrity of the olfactory system; instead, clinicians more commonly rely on behavioral tests of olfaction. It is possible that, for people, behavioral effects may prove to be more sensitive than nasal pathology in detecting toxicant-induced injury to the olfactory system. This dichotomy of responses between the species may reflect the relative importance of the sense of smell.

232

David C. Dorman and Melanie L. Foster

References 1. Dorman DC (2018) Olfactory system. In: McQueen CA (ed) Comprehensive toxicology, 3rd edn. Elsevier, New York, NY, pp 361–375 2. Harkema JR, Carey SA, Wagner JG (2006) The nose revisited: a brief review of the comparative structure, function, and toxicologic pathology of the nasal epithelium. Toxicol Pathol 34:252–269 3. Solbu TT, Holen T (2012) Aquaporin pathways and mucin secretion of Bowman’s glands might protect the olfactory mucosa. Chem Senses 37(1):35–46 4. Sarnat HB, Flores-Sarnat L (2017) Olfactory development, part 2: neuroanatomic maturation and dysgeneses. J Child Neurol 32:579–593 5. Badonnel K, Durieux D, Monnerie R, Gre´bert D, Salesse R, Caillol M, Baly B (2009) Leptin-sensitive OBP-expressing mucous cells in rat olfactory epithelium: a novel target for olfaction-nutrition crosstalk? Cell Tissue Res 338:53–66 6. Jeffrey AM, Iatropoulos MJ, Williams GM (2006) Nasal cytotoxic and carcinogenic activities of systemically distributed organic chemicals. Toxicol Pathol 34:827–852 7. Gross EA, Swenberg JA, Fields S, Popp JA (1982) Comparative morphometry of the nasal cavity in rats and mice. J Anat 135:83–88 8. Smith TD, Bhatnagar KP, Tuladhar P, Burrows AM (2004) Distribution of olfactory epithelium in the primate nasal cavity: are microsmia and macrosmia valid morphological concepts? Anat Rec A Discov Mol Cell Evol Biol 281:1173–1181 9. Craven BA, Paterson EG, Settles GS (2010) The fluid dynamics of canine olfaction: unique nasal airflow patterns as an explanation of macrosmia. J R Soc Interface 7:933–943 10. Rygg AD, Van Valkenburgh B, Craven BA (2017) The influence of sniffing on airflow and odorant deposition in the canine nasal cavity. Chem Senses 42:683–698 11. Beites CL, Kawauchi S, Crocker CE, Calof AL (2005) Identification and molecular regulation of neural stem cells in the olfactory epithelium. Exp Cell Res 306:309–316 12. Mackay-Sim A (2010) Stem cells and their niche in the adult olfactory mucosa. Arch Ital Biol 148:47–58 13. Schwob JE, Jang W, Holbrook EH, Lin B, Herrick DB, Peterson JN, Hewitt Coleman J (2017) Stem and progenitor cells of the

mammalian olfactory epithelium: taking poietic license. J Comp Neurol 525:1034–1054 14. Yu CR, Wu Y (2017) Regeneration and rewiring of rodent olfactory sensory neurons. Exp Neurol 287:395–408 15. McEwen DP, Jenkins PM, Martens JR (2008) Olfactory cilia: our direct neuronal connection to the external world. Curr Top Dev Biol 85:333–370 16. Pelosi P, Zhu J, Knoll W (2018) Odorantbinding proteins as sensing elements for odour monitoring. Sensors (Basel) 18(10): pii: E3248 17. Sun JS, Xiao S, Carlson JR (2018) The diverse small proteins called odorant-binding proteins. Open Biol 8:180208 18. Baker H, Grillo M, Margolis FL (1989) Biochemical and immunocytochemical characterization of olfactory marker protein in the rodent central nervous system. J Comp Neurol 285:246–261 19. Ressler KJ, Sullivan SL, Buck LB (1993) A zonal organization of odorant receptor gene expression in the olfactory epithelium. Cell 73:597–609 20. Sarnat HB, Flores-Sarnat L (2019) Development of the human olfactory system. Handb Clin Neurol 164:29–45 21. Zou DJ, Chesler A, Firestein S (2009) How the olfactory bulb got its glomeruli: a just so story? Nat Rev Neurosci 10:611–618 22. Lim DA, Alvarez-Buylla A (2016) The adult ventricular-subventricular zone (V-SVZ) and olfactory bulb (OB) neurogenesis. Cold Spring Harb Perspect Biol 8(5):pii: a018820 23. Heydel JM, Coelho A, Thiebaud N, Legendre A, Le Bon AM, Faure P, Neiers F, Artur Y, Golebiowski J, Briand L (2013) Odorant-binding proteins and xenobiotic metabolizing enzymes: implications in olfactory perireceptor events. Anat Rec (Hoboken) 296:1333–1345 24. Grabe V, Sachse S (2018) Fundamental principles of the olfactory code. Biosystems 164:94–101 25. Genter MB (2004) Update on olfactory mucosal metabolic enzymes: age-related changes and N-acetyltransferase activities. J Biochem Mol Toxicol 18:239–244 26. Watelet JB, Strolin-Benedetti M, Whomsley R (2009) Defence mechanisms of olfactory neuro-epithelium: mucosa regeneration, metabolising enzymes and transporters. B-ENT Suppl 13:21–37

Assessment of Olfactory Toxicity in Rodents 27. Ling G, Gu J, Genter MB, Zhuo X, Ding X (2004) Regulation of cytochrome P450 gene expression in the olfactory mucosa. Chem Biol Interact 147:247–258 28. Dahl AR, Hadley WM (1991) Nasal cavity enzymes involved in xenobiotic metabolism: effects on the toxicity of inhalants. Crit Rev Toxicol 21:345–372 29. Reed CJ (1993) Drug metabolism in the nasal cavity: relevance to toxicology. Drug Metab Rev 25:173–205 30. Burckhardt G (2012) Drug transport by organic anion transporters (OATs). Pharmacol Ther 136:106–130 31. Thiebaud N, Menetrier F, Belloir C, Minn AL, Neiers F, Artur Y, Le Bon AM, Heydel JM (2011) Expression and differential localization of xenobiotic transporters in the rat olfactory neuro-epithelium. Neurosci Lett 505:180–185 32. Lunde K, Egelandsdal B, Skuterud E, Mainland JD, Lea T, Hersleth M, Matsunami H (2012) Genetic variation of an odorant receptor OR7D4 and sensory perception of cooked meat containing androstenone. PLoS One 7 (5):e35259 33. Boesveldt S, Postma EM, Boak D, WelgeLuessen A, Scho¨pf V, Mainland JD, Martens J, Ngai J, Duffy VB (2017) Anosmia—a clinical review. Chem Senses 42:513–523 34. Doty RL (2018) Age-related deficits in taste and smell. Otolaryngol Clin N Am 51:815–825 35. Genter MB, Doty RL (2019) Toxic exposures and the senses of taste and smell. Handb Clin Neurol 164:389–408 36. Coronas-Sa´mano G, Portillo W, Beltra´n Campos V, Medina-Aguirre GI, Paredes RG, Diaz-Cintra S (2014) Deficits in odor-guided behaviors in the transgenic 3xTg-AD female mouse model of Alzheimer’s disease. Brain Res 1572:18–25 37. Bermu´dez ML, Seroogy KB, Genter MB (2019) Evaluation of carnosine intervention in the Thy1-aSyn mouse model of Parkinson’s disease. Neuroscience 411:270–278 38. Croy I, Nordin S, Hummel T (2014) Olfactory disorders and quality of life—an updated review. Chem Senses 39:185–194 39. Gobba F (2006) Olfactory toxicity: long-term effects of occupational exposures. Int Arch Occup Environ Health 79:322–331 40. Heidari F, Karimi E, Firouzifar M, Khamushian P, Ansari R, Mohammadi Ardehali M, Heidari F (2020) Anosmia as a

233

prominent symptom of COVID-19 infection. Rhinology 58:302–303 41. Hopkins C, Surda P, Kumar N (2020) Presentation of new onset anosmia during the COVID-19 pandemic. Rhinology 58:295–298 42. Kaye R, Chang CWD, Kazahaya K, Brereton J, Denneny JC 3rd (2020) COVID-19 anosmia reporting tool: initial findings. Otolaryngol Head Neck Surg 163:132–134 43. Klopfenstein T, Kadiane-Oussou NJ, Toko L, Royer PY, Lepiller Q, Gendrin V, Zayet S (2020) Features of anosmia in COVID-19. Med Mal Infect 50(5):436–439 44. Wrobel BB, Leopold DA (2004) Clinical assessment of patients with smell and taste disorders. Otolaryngol Clin N Am 37:1127–1142 45. Doty RL, Shaman P, Kimmelman CP, Dann MS (1984) University of Pennsylvania Smell Identification Test: a rapid quantitative olfactory function test for the clinic. Laryngoscope 94:176–178 46. Doty RL (2018) Measurement of chemosensory function. World J Otorhinolaryngol Head Neck Surg 4:11–28 47. Radil T, Wysocki CJ (1998) Spatiotemporal masking in pure olfaction. Ann N Y Acad Sci 855:641–644 48. Tremblay C, Frasnelli J (2018) Olfactory and trigeminal systems interact in the periphery. Chem Senses 43:611–616 49. Owens JG, James RA, Moss OR, Morgan KT, Bowman JR, Struve MF, Dorman DC (1996) Design and evaluation of an olfactometer for the assessment of 3-methylindole-induced hyposmia. Fundam Appl Toxicol 33:60–70 50. Foster ML, Rao DB, Francher T, Traver S, Dorman DC (2018) Olfactory toxicity in rats following manganese chloride nasal instillation: a pilot study. Neurotoxicology 64:284–290 51. Chamanza R, Wright JA (2015) A review of the comparative anatomy, histology, physiology and pathology of the nasal cavity of rats, mice, dogs and non-human primates. Relevance to inhalation toxicology and human health risk assessment. J Comp Pathol 153:287–314 52. Morgan KT, Monticello TM (1990) Airflow, gas deposition, and lesion distribution in the nasal passages. Environ Health Perspect 85:209–218 53. Maronpot RR (1990) Pathology Working Group review of selected upper respiratory

234

David C. Dorman and Melanie L. Foster

tract lesions in rats and mice. Environ Health Perspect 85:331–352 54. Monticello TM, Morgan KT, Uraih L (1990) Nonneoplastic nasal lesions in rats and mice. Environ Health Perspect 85:249–274 55. Mery S, Gross EA, Joyner DR, Godo M, Morgan KT (1994) Nasal diagrams: a tool for recording the distribution of nasal lesions in rats and mice. Toxicol Pathol 22:353–372 56. Morgan KT (1991) Approaches to the identification and recording of nasal lesions in toxicology studies. Toxicol Pathol 19:337–351 57. Dorman DC (2019) Use of nasal pathology in the derivation of inhalation toxicity values for hydrogen sulfide. Toxicol Pathol 47:1043–1048 58. Organization for Economic Co-operation and Development (2009). Draft OECD guidance document on histopathology for inhalation toxicity studies, supporting TG 412 (subacute inhalation toxicity: 28-Day) and TG 413 (subchronic inhalation toxicity: 90-Day). http://www.oecd.org.prox.lib. ncsu.edu/chemicalsafety/testing/ 43062801.pdf. Accessed 4 Sep 2020 59. Gross EA, Patterson DL, Morgan KT (1987) Effects of acute and chronic dimethylamine exposure on the nasal mucociliary apparatus of F-344 rats. Toxicol Appl Pharmacol 90:359–376 60. Wolf DC, Morgan KT, Gross EA, Barrow C, Moss OR, James RA, Popp JA (1995) Two-year inhalation exposure of female and male B6C3F1 mice and F344 rats to chlorine gas induces lesions confined to the nose. Fundam Appl Toxicol 24:111–131 61. Morrison EE, Costanzo RM (1992) Morphology of olfactory epithelium in humans and other vertebrates. Microsc Res Tech 23:49–61 62. Gao L, Yang ST, Li S, Meng Y, Wang H, Lei H (2013) Acute toxicity of zinc oxide nanoparticles to the rat olfactory system after intranasal instillation. J Appl Toxicol 33:1079–1088 63. Wiethoff AJ, Harkema JR, Koretsky AP, Brown WE (2001) Identification of mucosal injury in the murine nasal airways by magnetic resonance imaging: site-specific lesions induced by 3-methylindole. Toxicol Appl Pharmacol 175:68–75 64. Kimbell JS, Subramaniam RP (2001) Use of computational fluid dynamics models for dosimetry of inhaled gases in the nasal passages. Inhal Toxicol 13:325–334 65. Moulin FJ, Brenneman KA, Kimbell JS, Dorman DC (2002) Predicted regional flux of

hydrogen sulfide correlates with distribution of nasal olfactory lesions in rats. Toxicol Sci 66:7–15 66. Schroeter JD, Kimbell JS, Gross EA, Willson GA, Dorman DC, Tan YM, Clewell HJ 3rd (2008) Application of physiological computational fluid dynamics models to predict interspecies nasal dosimetry of inhaled acrolein. Inhal Toxicol 20:227–243 67. Schellinck H (2018) Measuring olfactory processes in Mus musculus. Behav Process 155:19–25 68. Wackermannova´ M, Pinc L, Jebavy´ L (2016) Olfactory sensitivity in mammalian species. Physiol Res 65:369–390 69. Moser VC (2011) Functional assays for neurotoxicity testing. Toxicol Pathol 39:36–45 70. Gregory EH, Pfaff DW (1971) Development of olfactory-guided behavior in infant rats. Physiol Behav 6:573–576 71. Kaneko-Goto T, Sato Y, Katada S, Kinameri E, Yoshihara S, Nishiyori A, Kimura M, Fujita H, Touhara K, Reed RR, Yoshihara Y (2013) Goofy coordinates the acuity of olfactory signaling. J Neurosci 33:12,987–12,996a 72. Staples LG, McGregor IS, Apfelbach R, Hunt GE (2008) Cat odor, but not trimethylthiazoline (fox odor), activates accessory olfactory and defense-related brain regions in rats. Neuroscience 151:937–947 73. Yang M, Crawley JN (2009) Simple behavioral assessment of mouse olfaction. Curr Protoc Neurosci Chapter 8:Unit 8.24 74. Hao S, Yu F, Yan A, Zhang Y, Han J, Jiang X (2012) In utero and lactational lanthanum exposure induces olfactory dysfunction associated with downregulation of βIII-tubulin and olfactory marker protein in young rats. Biol Trace Elem Res 148:383–391 75. Hsieh H, Horwath MC, Genter MB (2017) Zinc gluconate toxicity in wild-type vs. MT1/ 2-deficient mice. Neurotoxicology 58:130–136 76. Hurtt ME, Thomas DA, Working PK, Monticello TM, Morgan KT (1988) Degeneration and regeneration of the olfactory epithelium following inhalation exposure to methyl bromide: pathology, cell kinetics, and olfactory function. Toxicol Appl Pharmacol 94:311–328 77. Liu Y, Sun JD, Song LK, Li J, Chu S, Yuan Y, Chen N (2015) Environment-contact administration of rotenone: a new rodent model of Parkinson’s disease. Behav Brain Res 294:149–161

Assessment of Olfactory Toxicity in Rodents 78. Lehmkuhl AM, Dirr ER, Fleming SM (2014) Olfactory assays for mouse models of neurodegenerative disease. J Vis Exp 90:e51804 79. Oummadi A, Meyer-Dilhet G, Be´ry A, Aubert A, Barone P, Mortaud S, Guillemin GJ, Menuet A, Laugeray A (2020) 3Rs-based optimization of mice behavioral testing: the habituation/dishabituation olfactory test. J Neurosci Methods 332:108550 80. Campolongo M, Kazlauskas N, Falasco G, Urrutia L, Salgueiro N, Ho¨cht C, Depino AM (2018) Sociability deficits after prenatal exposure to valproic acid are rescued by early social enrichment. Mol Autism 9:36 81. De Felice A, Venerosi A, Ricceri L, Sabbioni M, Scattoni ML, Chiarotti F, Calamandrei G (2014) Sex-dimorphic effects of gestational exposure to the organophosphate insecticide chlorpyrifos on social investigation in mice. Neurotoxicol Teratol 46:32–39 82. Tarland E, Brosda J (2018) Male rats treated with subchronic PCP show intact olfaction and enhanced interest for a social odour in the olfactory habituation/dishabituation test. Behav Brain Res 345:13–20 83. Czarnecki LA, Moberly AH, Rubinstein T, Turkel DJ, Pottackal J, McGann JP (2011) In vivo visualization of olfactory pathophysiology induced by intranasal cadmium instillation in mice. Neurotoxicology 32:441–449 84. Ferry AT, Lu XC, Price JL (2000) Effects of excitotoxic lesions in the ventral striatopallidal--thalamocortical pathway on odor reversal learning: inability to extinguish an incorrect response. Exp Brain Res 131:320–335 85. Sokolic L, Long LE, Hunt GE, Arnold JC, McGregor IS (2011) Disruptive effects of the prototypical cannabinoid Δ9-tetrahydrocannabinol and the fatty acid amide inhibitor URB-597 on Go/No-Go auditory discrimination performance and olfactory reversal learning in rats. Behav Pharmacol 22:191–202 86. Arbuckle EP, Smith GD, Gomez MC, Lugo JN (2015) Testing for odor discrimination and habituation in mice. J Vis Exp 99:e52615 87. Liu G, Patel JM, Tepe B, McClard CK, Swanson J, Quast KB, Arenkiel BR (2018) An objective and reproducible test of olfactory learning and discrimination in mice. J Vis Exp 133:e57142 88. Clevenger AC, Restrepo D (2006) Evaluation of the validity of a maximum likelihood adaptive staircase procedure for measurement of olfactory detection threshold in mice. Chem Senses 31:9–26

235

89. Walker JC, O’Connell RJ (1986) Computerized odor psychophysical testing in mice. Chem Senses 11(4):439–453 90. Youngentob SL, Schwob JE, Sheehe PR, Youngentob LM (1997) Odorant threshold following methyl bromide-induced lesions of the olfactory epithelium. Physiol Behav 62:1241–1252 91. Chen X, Xia Z, Storm DR (2013) Electroolfactogram (EOG) recording in the mouse main olfactory epithelium. Bio Protoc 3(11): pii: e789 92. Coppola DM, Waggener CT, Radwani SM, Brooks DA (2013) An electroolfactogram study of odor response patterns from the mouse olfactory epithelium with reference to receptor zones and odor sorptiveness. J Neurophysiol 109:2179–2191 93. Cygnar KD, Stephan AB, Zhao H (2010) Analyzing responses of mouse olfactory sensory neurons using the air-phase electroolfactogram recording. J Vis Exp 37:pii: 1850 94. Pinato G, Rievaj J, Pifferi S, Dibattista M, Masten L, Menini A (2008) Electroolfactogram responses from organotypic cultures of the olfactory epithelium from postnatal mice. Chem Senses 33:397–404 95. Faure F, Da Silva SV, Jakob I, Pasquis B, Sicard G (2010) Peripheral olfactory sensitivity in rodents after treatment with docetaxel. Laryngoscope 120:690–697 96. Thiebaud N, Veloso Da Silva S, Jakob I, Sicard G, Chevalier J, Me´ne´trier F, Berdeaux O, Artur Y, Heydel JM, Le Bon AM (2013) Odorant metabolism catalyzed by olfactory mucosal enzymes influences peripheral olfactory responses in rats. PLoS One 8(3):e59547 97. Aydemir O (2017) Olfactory recognition based on EEG gamma-band activity. Neural Comput 29:1667–1680 98. Lorig TS (2000) The application of electroencephalographic techniques to the study of human olfaction: a review and tutorial. Int J Psychophysiol 36:91–104 99. Walla P, Hufnagl B, Lehrner J, Mayer D, Lindinger G, Imhof H, Deecke L, Lang W (2003) Olfaction and face encoding in humans: a magnetoencephalographic study. Brain Res Cogn Brain Res 15:105–115 100. Gudziol H, Guntinas-Lichius O (2019) Electrophysiologic assessment of olfactory and gustatory function. Handb Clin Neurol 164:247–262

236

David C. Dorman and Melanie L. Foster

101. Masaoka Y, Harding IH, Koiwa N, Yoshida M, Harrison BJ, Lorenzetti V, Ida M, Izumizaki M, Pantelis C, Homma I (2014) The neural cascade of olfactory processing: a combined fMRI-EEG study. Respir Physiol Neurobiol 204:71–77 102. Gervais R, Buonviso N, Martin C, Ravel N (2007) What do electrophysiological studies tell us about processing at the olfactory bulb level? J Physiol Paris 101:40–45 103. Zhang B, Zhuang L, Qin Z, Wei X, Yuan Q, Qin C, Wang P (2018) A wearable system for olfactory electrophysiological recording and animal motion control. J Neurosci Methods 307:221–229 104. Benveniste H, Lazebnik Y, Volkow ND (2017) Seeing how we smell. J Clin Invest 127:447–449 105. Kim YK, Hong SL, Yoon EJ, Kim SE, Kim JW (2012) Central presentation of postviral olfactory loss evaluated by positron emission tomography scan: a pilot study. Am J Rhinol Allergy 26:204–208 106. Risacher SL, Tallman EF, West JD, Yoder KK, Hutchins GD, Fletcher JW, Gao S, Kareken DA, Farlow MR, Apostolova LG, Saykin AJ (2017) Olfactory identification in subjective cognitive decline and mild cognitive impairment: association with tau but not amyloid positron emission tomography. Alzheimers Dement (Amst) 9:57–66 107. Han P, Zang Y, Akshita J, Hummel T (2019) Magnetic resonance imaging of human olfactory dysfunction. Brain Topogr 32:987–997 108. Levy LM, Henkin RI, Hutter A, Lin CS, Martins D, Schellinger D (1997) Functional MRI of human olfaction. J Comput Assist Tomogr 21:849–856 109. Moon WJ, Park M, Hwang M, Kim JK (2018) Functional MRI as an objective measure of olfaction deficit in patients with traumatic anosmia. AJNR Am J Neuroradiol 39:2320–2325 110. Chuang KH, Lee JH, Silva AC, Belluscio L, Koretsky AP (2009) Manganese enhanced MRI reveals functional circuitry in response to odorant stimuli. NeuroImage 44:363–372 111. Han Z, Chen W, Chen X, Zhang K, Tong C, Zhang X, Li CT, Liang Z (2019) Awake and behaving mouse fMRI during Go/No-Go task. NeuroImage 188:733–742 112. Lehallier B, Coureaud G, Maurin Y, Bonny JM (2012) Effects of manganese injected into rat nostrils: implications for in vivo functional study of olfaction using MEMRI. Magn Reson Imaging 30:62–69

113. Litaudon P, Bouillot C, Zimmer L, Costes N, Ravel N (2017) Activity in the rat olfactory cortex is correlated with behavioral response to odor: a microPET study. Brain Struct Funct 222:577–586 114. Muir ER, Biju KC, Cong L, Rogers WE, Torres Hernandez E, Duong TQ, Clark RA (2019) Functional MRI of the mouse olfactory system. Neurosci Lett 704:57–61 115. Van de Bittner GC, Riley MM, Cao L, Ehses J, Herrick SP, Ricq EL, Wey HY, O’Neill MJ, Ahmed Z, Murray TK, Smith JE, Wang C, Schroeder FA, Albers MW, Hooker JM (2017) Nasal neuron PET imaging quantifies neuron generation and degeneration. J Clin Invest 127:681–694 116. Courtiol E, Neiman M, Fleming G, Teixeira CM, Wilson DA (2019) A specific olfactory cortico-thalamic pathway contributing to sampling performance during odor reversal learning. Brain Struct Funct 224:961–971 117. Ogg MC, Ross JM, Bendahmane M, Fletcher ML (2018) Olfactory bulb acetylcholine release dishabituates odor responses and reinstates odor investigation. Nat Commun 9:1868 118. Wilson CD, Serrano GO, Koulakov AA, Rinberg D (2017) A primacy code for odor identity. Nat Commun 8(1):1477 119. Kass MD, Czarnecki LA, McGann JP (2018) Stable olfactory sensory neuron in vivo physiology during normal aging. Neurobiol Aging 69:33–37 120. Lee AC, Tian H, Grosmaitre X, Ma M (2009) Expression patterns of odorant receptors and response properties of olfactory sensory neurons in aged mice. Chem Senses 34:695–703 121. Kunkhyen T, Perez E, Bass M, Coyne A, Baum MJ, Cherry JA (2018) Gonadal hormones, but not sex, affect the acquisition and maintenance of a Go/No-Go odor discrimination task in mice. Horm Behav 100:12–19 122. Kass MD, Czarnecki LA, Moberly AH, McGann JP (2017) Differences in peripheral sensory input to the olfactory bulb between male and female mice. Sci Rep 7:45851 123. van der Linden C, Jakob S, Gupta P, Dulac C, Santoro SW (2018) Sex separation induces differences in the olfactory sensory receptor repertoires of male and female mice. Nat Commun 9:5081 124. Ueno H, Shimada A, Suemitsu S, Murakami S, Kitamura N, Wani K, Takahashi Y, Matsumoto Y, Okamoto M, Fujiwara Y, Ishihara T (2019) Comprehensive

Assessment of Olfactory Toxicity in Rodents behavioral study of the effects of vanillin inhalation in mice. Biomed Pharmacother 115:108879 125. Lo¨tsch J, Ultsch A, Hummel T (2016) How many and which odor identification items are needed to establish normal olfactory function? Chem Senses 41:339–344 126. Persaud KC (2013) Engineering aspects of olfaction. In: Persaud KC, Marco S, Gutie´rrez-Ga´lvez A (eds) Neuromorphic olfaction.

237

CRC Press/Taylor & Francis, Boca Raton, FL; Chapter 1. Frontiers in neuroengineering 127. Evans JE, Miller ML, Andringa A, Hastings L (1995) Behavioral, histological, and neurochemical effects of nickel (II) on the rat olfactory system. Toxicol Appl Pharmacol 130:209–220 128. Lu XC, Slotnick BM (1998) Olfaction in rats with extensive lesions of the olfactory bulbs: implications for odor coding. Neuroscience 84:849–866

Chapter 11 Assessment of Neurotoxicant-Induced Changes in Behavior: Issues Related to Interpretation of Outcomes and Experimental Design Deborah A. Cory-Slechta, Katherine Harvey, and Marissa Sobolewski Abstract Behavioral assessment is a critical component of neurotoxicological research as the consequences of neurotoxicant exposure frequently include changes in behavior. Behavior encompasses multiple domains, and numerous behavioral testing paradigms, ranging from simple to complex, are available for assessment within each such domain and described in the literature. Many laboratories adapt the simplest procedures, based on the assumption that these procedures would be both simple to implement and to interpret. Such assumptions fail to recognize that behavior is actually quite complex. For example, it is critical to consider potential confounding effects (stress, nutritional state, motivation, motor endurance, fear, pain sensitivity) inherent to common behavioral tests, i.e., to understand behavioral mechanisms of an observed behavioral change, because they may shift the interpretation of a change in behavior, particularly for cognitive/ executive functions, such as learning. Additionally, this chapter addresses issues related to the use and misuse of test batteries and the influence of the behavioral history resulting from use of a test battery that may interact with the toxicant treatment. It also emphasizes the critical need to recognize that behavioral testing itself modifies the brain; hence, attempting to define biological mechanisms of a neurotoxicant from brains of organisms that have behavioral experience may prove misleading, as brain changes will reflect the effects of both the toxicant exposure and the behavioral experience. Finally, the chapter emphasizes the advantages of using the same behavioral paradigms across species, with appropriate confounds measured and ethologically relevant modifications made, to enhance translation. Key words Behavioral experience, Behavioral history, Behavioral mechanisms, Neurotransmitters, Lead, Prenatal stress, Translation

1

Introduction and Purpose The brain is comprised of the dynamic and highly interactive activity of interdependent cells, networks and systems, and an organism’s behavior reflects the functional output of these systems (Fig. 1). Additionally, the brain interacts with and controls functions of peripheral organs and systems. It is critical to recognize that behavior does not originate from molecular events within the central nervous system, but is ultimately a response to environmental

Jordi Llorens and Marta Barenys (eds.), Experimental Neurotoxicology Methods, Neuromethods, vol. 172, https://doi.org/10.1007/978-1-0716-1637-6_11, © Springer Science+Business Media, LLC, part of Springer Nature 2021

239

240

Deborah A. Cory-Slechta et al.

Fig. 1 Behavior originates in the environment while physiology creates the framework within which the environment influences behavior. Consequences of behavior can feed back to modify the physiological framework

stimuli, i.e., behavior is context-dependent. Different molecular environments can change the probability that an organism will respond in a particular way to a particular environmental stimulus. But it is simplistic to consider that gene x environment interactions are constants, e.g., that simply activating gene A or hormone B will invariably invoke behavior Y across all environmental contexts in all individuals [1]. For example, an evolving understanding demonstrates that the relationship between increased testosterone and aggressive behavior is not linear or invariant, but, rather, is dependent on the type of aggression, on genetic sex, on parental status, on age, and more [2, 3]. However, providing environmental context improves our understanding of when such relationships will exist and when they do not. A common adage has been that genes + environment ¼ phenotype. However, to define it more completely, organismal physiology (and its related subcomponents) + organismal environment (with its behavioral history and context) ¼ behavior (Fig. 1). Associations of biochemical effects with behavior can be altered based on variations in nuclear, cellular, organ, and physiological environments, and these environments are molded and shaped by previous rewards, behavioral histories, and social interactions in a highly interactive and dynamic capacity across time in behaving organisms. Within that framework, interpretations of behavioral toxicity can only be derived from an understanding of an organism’s previous behavioral history, social rank/relationships, stimulus conditioning history, context-dependency, and/or the deprivation of these experiences. Further, many behavioral test methods designed to evaluate behavioral domains such as learning, particularly those

Assessment of Neurotoxicant-Induced Changes in Behavior: Issues Related. . .

241

based on simple assays, actually rely on a behavioral response that encompasses other behavioral domains, i.e., motor and sensory function, which if altered by the toxicant, will itself alter learning. Without an appreciation of this bidirectional influence, we risk missing the potential confounding physiological and environmental factors in our outcomes, and our predictive validity and ability to reproduce associations will be limited and potentially misleading, as this chapter seeks to further elaborate through discussions of: (1) the importance of confounding environmental and/or physiological influences on “behavioral mechanisms” of action and (2) the influences of prior behavioral experience/history on neurotoxicology studies. In addition to understanding the context of the origins of an organism’s behavior, consideration of the dynamic temporal relationship between a biochemical response and a behavioral response is critical. Behavioral responses provoked by a stimulus can occur extremely rapidly, i.e., sometimes within milliseconds of a stimulus. Consider, e.g., the rapidity of removing one’s hand from a sudden fire. Fire is clearly the environmental context for the response, and such rapidity and the plasticity of behavior are critical in this example to prevent burns. This response reflects the interaction between environment and physiological functions, such as neuronal myelination, synaptic connectivity, and activation of endocrine stress responses. In contrast, biological changes, such as in gene expression and protein synthesis, require a far greater time frame to process and would, if such factors actually controlled behavior, already have resulted in burns. Instead the association between transcriptional shifts and/or epigenetic reprogramming and an environmental stimulus is often temporally delayed. In fact, the behavioral response to a stimulus alters an organism’s physiology through these mechanisms to provide a “memory” of this event that then defines the parameters of future responses, including rapidity and plasticity, by molding cellular signaling and neuronal function. As this implies, genes and the nuclear environment, with their influences on physiology and morphology, act to create a framework or parameters within which the environment acts to shape the behavior of an individual. Behavioral responses can feed back throughout the organism to alter this biological framework at multiple levels. As characterized by Skinner: “We shall know the precise neurological conditions which immediately precede, say, the response, “No, thank you.” These events in turn will be found to be preceded by other neurological events, and these in turn by others. This series will lead us back to events outside the nervous system and, eventually, outside the organism” [4]. This is the framework requisite for interpretation of behavioral toxicity, particularly for ostensible changes in executive functions. Behavior is a critical component both of the understanding of the neurotoxic

242

Deborah A. Cory-Slechta et al.

consequences of environmental chemical exposures, as well as to advancing our understanding of brain function, as it is often one of the most sensitive and translational endpoints. Both simple and complex methods exist to assess virtually all domains of behavioral function. Reported descriptions and instructions for implementation of a sizeable array of specific behavioral procedures for carrying out such assessments are already widely available for almost every behavioral domain [5] and include parametric considerations related to the species being tested. Probably more so than in any other area, the scientific community appears to consider it relatively simple to implement behavioral paradigms into laboratories even without any background or expertise. This tendency was presaged by Skinner [4] who stated that “Actually there is no subject matter with which we could be better acquainted, for we are always in the presence of at least one behaving organism.” However, as he also went on to note: “But this familiarity is something of a disadvantage, for it means that we have probably jumped to conclusions which will not be supported by the cautious methods of science.” This quote seems particularly relevant today given the tendency of many laboratories to utilize the simplest behavioral methods in lieu of more complex techniques, likely given the greater rapidity and lowered costs with which they can be implemented. However, as has become clear through years of the experimental analysis of behavior, the less complicated the procedure, the more difficult to interpret the nature of an observed behavioral change and whether the observed change actually or specifically reflects the behavioral domain ostensibly being tested in a given behavioral procedure, or whether alternate mechanisms are operative [6]. For example, in the use of a supposed “learning” paradigm, a behavioral change in response to a toxicant may not reflect a learning deficit at all, but instead a change in motivation and/or motor function and/or sensory capacities, all of which are components of the designated response/procedure being used to evaluate learning. Thus, it is inappropriate to presume that because a behavioral procedure is stated to measure learning, that any observed behavioral change in that procedure in response to a toxicant exposure is necessarily a learning deficit. While the scientific community prefers simplicity of explanation and unqualified, broad generalizations, this chapter focuses on the potential problems that can arise when behavioral context is ignored as it relates to: the interpretation of outcomes from various behavioral procedures, to appropriate experimental designs, as well as to reliability and reproducibility of effects. After discussing the origins of behavior, it endeavors to underscore the importance of recognizing how different physiological and environmental contexts alter behavioral mechanisms of action, and how these alterations influence behavior, particularly in relation to the least

Assessment of Neurotoxicant-Induced Changes in Behavior: Issues Related. . .

243

complicated behavioral assays that are most widely used. Secondly, it discusses the potential limitations and confounds to be considered when using a series of behavioral tests, i.e., a behavioral battery, in the same subjects. Thirdly, it re-introduces the associated topic of “behavioral history,” i.e., the influence of behavioral experience itself on subsequent behavior, drug effects, etc. In addition, it emphasizes the potential limitations and confounds introduced by the fact that behavioral assessment itself introduces changes in brain, indeed, across all scales, as this has significant consequences for experimental design and interpretation, particularly as related to biological mechanisms of action of a neurotoxicant. Finally, it emphasizes the utility of the use of the same behavioral paradigms across species to minimize difficulties of extrapolation to human subjects.

2 Origins of Behavior: Evolution, Phylogeny, Development, and Individual Experience The behavioral repertoire ranges from unlearned to learned behaviors. The overlaps across species in characteristic behavioral repertoires are a function of shared phylogenetic ancestry and evolutionary adaptive significance [7, 3]. Unlearned behaviors include unconditioned reflexes (URs) and some fixed action patterns. However, some unlearned behaviors, particularly some fixed action patterns, actually have the capacity to be modified [8]. The bounds of learned behavioral flexibility reflect cumulative developmental and individual behavioral history. Additionally, explanations for the origins of learned behaviors overlap, as throughout evolution, primates (and rodents) have been evolutionarily selected for increased behavioral and phenotypic plasticity, increasing the role of our developmental environments and individual learning experiences to produce wide behavioral complexity. Two processes underlie the origins of learned behavior across species, specifically Pavlovian conditioning and operant conditioning. Pavlovian conditioning (Fig. 2) builds upon unconditioned reflexes (UR) built into the organism, which are elicited by an unconditioned environmental stimulus, e.g., a light shining into the eye (US; unconditioned stimulus) elicits pupillary dilation (unconditioned response, UR) which can then serve as the basis for the establishment of new conditioned reflexes. In that second-order conditioning process, a new, initially neutral (from the organism’s perspective) stimulus is repeatedly paired with the unconditioned stimulus, with the neutral stimulus preceding the unconditioned stimulus. With multiple pairings, the neutral stimulus itself acquires conditioning properties, i.e., becomes a conditioned stimulus (CS) such that it alone can now elicit the

244

Deborah A. Cory-Slechta et al.

Fig. 2 Schematic depiction of respondent (left column) and operant (right column) conditioning paradigms. CS Conditioned stimulus, US Unconditioned stimulus; SR Reinforcer

unconditioned response, now defined as a conditioned reflex or conditioned response (CR). The classic example has been the repeated pairing of a tone with food powder such that the tone itself comes to elicit a conditioned salivation response in dogs. More complex and higher order Pavlovian conditioning can also occur, such as when another neutral stimulus is paired with a conditioned stimulus eliciting a conditioned response; with repeated pairing that neutral stimulus will also acquire conditioned stimulus properties. In the absence of intermittent re-pairing, conditioned stimuli lose their ability to elicit conditioned responses. The basis of operant conditioning (Fig. 2) is voluntary, not elicited, behavior. In operant conditioning, a voluntary response is followed by some environmental stimulus (consequence). If that stimulus is rewarding (reinforcing), the frequency of the response it followed will increase. A reinforcing stimulus may be a positive reinforcer (its presentation is reinforcing, e.g., delivery of a food reward following completion of a designated response) or a negative reinforcer (its removal is reinforcing, e.g., termination of an ongoing shock following completion of a designated response). If the reinforcing stimulus is discontinued, the frequency of the response will decline, a process known as extinction. If the environmental stimulus that follows a response is punishing (aversive stimulus), the frequency of the response it follows will decline. Initially, neutral stimuli can acquire reinforcing or aversive properties through pairing with other unconditioned (e.g., primary reinforcers) or with conditioned (secondary reinforcers) stimuli that have acquired such properties. Indeed, some such stimuli, i.e., generalized reinforcers, become associated with multiple other

Assessment of Neurotoxicant-Induced Changes in Behavior: Issues Related. . .

245

reinforcing stimuli through such pairings, such as in the case of money or attention. As an extensive experimental literature has documented over the years, these procedures and principles are operative across species.

3

Recognizing the Importance of Behavioral Mechanisms of Action As alluded to above, behavior is a reflection of the environment and of experience. While we often consider biological mechanisms of action, behavioral mechanisms of action and their potential influence on behavior, are generally not fully considered despite their critical role in interpretation of behavioral changes. These interdependent mechanisms include the antecedent stimulus environment, which can include interoceptive and/or exteroceptive stimuli, the nature of the designated response, and the ultimate consequences of the response, all of which ultimately influence outcome, as shown in Fig. 3. Antecedent conditions are critical, with environmental stimuli providing significant and crucial information as to the probability of how or whether a response will be reinforced; such stimuli are deemed discriminative stimuli. For example, if one sees that there are no cookies inside of a glass cookie jar, the odds that reaching into the jar would be reinforced are zero, likely then precluding emission of the response. Antecedent conditions likewise influence motivation, i.e., if you haven’t eaten for 24 h, food is highly likely to serve as a reinforcer, whereas for someone with a stomach flu (interoceptive stimulus events), this would not be the case. A red traffic light signals that braking behavior will be reinforced, while continuing to drive through it could be punished. Thus, environmental antecedents are critical in influencing the likelihood of a response. Notably, behavior under very strong control by the antecedent stimuli can also be resistant to effects of drugs and toxicants. For example, the effects of d-amphetamine were found to be significantly diminished in a context where a specific stimulus was presented when pigeons had reached the designated number of key pecking responses required such that a response on a different key would now produce food reward relative to the same behavioral procedure but without the signaling stimulus [9, 10]. A study that directly compared a nose-poke vs. a bar press response in mice found that administration of pentobarbital or methocarbamol (a muscle relaxant) prior to the session resulted in greater reduction of lever pressing responses than of nose-pokes [11], underscoring the importance of the topography of the chosen response. In another such report, acquisition of avoidance behavior of rats was found to be faster and to exhibit sharper temporal discrimination in a shuttle-box apparatus requiring a jumping response rather than in a Skinner box in which lever pressing served

246

Deborah A. Cory-Slechta et al.

Fig. 3 Behavioral mechanisms of action: Behavior occurs in a context. The likelihood of a response occurring depends upon antecedent conditions, e.g., motiviation, stimuli signaling likelihood of reward. Response topography and its relevance to the species being studied can influence the rate at which a designated response may occur, while consequences of the response determine its future probability of reoccurrence

as the operant response [12]. Hence, ease and species appropriateness of the nature and parameters of the designated response can increase reward densities and learning opportunities. The third component of this behavioral sequence is the nature of the consequence that follows the response. A consequence that follows a response is only defined as a reinforcer if it increases the probability of the response that it follows, or only a punisher if it decreases the frequency of behavior. As a corollary, it is important to recognize that the reinforcing or aversive properties of an environmental stimulus are not invariant, they are context-dependent. For example, as alluded to above, food may not serve as a reinforcer for an organism that is already sated, i.e., the behavioral mechanism is related to the establishing operation of pre-session feeding or restriction of food. In addition to the examples cited above, a counterintuitive example is that shock delivery itself, normally considered an inherently aversive stimulus, has been shown under certain environmental and behavioral history conditions to serve as a positive reinforcer [13]. Importantly, the timing of reward delivery is critical. In fact, delay of reward has long been known to hinder learning, and it can end up rewarding the wrong behavior, resulting in the increased frequency of superstitious behaviors [14–16].

4 Consideration of Behavioral Mechanisms in Interpreting the Specificity of the Behavioral Produced for Measuring a Particular Behavioral Function Recognition of the role of the three-term contingency shown in Fig. 3 has particular significance for interpretation of the outcome of neurotoxicant exposure on behavior. As noted, behavior has been characterized into different domains, such as motor function,

Assessment of Neurotoxicant-Induced Changes in Behavior: Issues Related. . .

247

Fig. 4 A wide array of techniques, ranging from very simple to far more complex are available for measurement of virtually every behavioral domain. However, the simpler the technique, the more difficult it is to determine the specificity of the observed behavior

sensory behavior, and executive responses as well as social behaviors. Multiple behavioral procedures have been developed for assessment of each of these domains and, within each such domain, behavioral testing procedures can range from very uncomplicated assays to more complex approaches in terms of, e.g., equipment, behavioral training requirements, ease of implementation, etc. The trend in behavioral toxicological research, and in neuroscience in general, has predominantly been to utilize uncomplicated behavioral assays as they are easier to implement, provide data more rapidly and are not perceived to require significant behavioral expertise within the laboratory. In general, however, the simpler the behavioral assay used, the more difficult it can be to interpret the outcome (Fig. 4). Consider simple paradigms used to measure “learning.” In a technique such as a Y maze or a T maze, the subject learns to turn to the side of the maze that contains a presumed reinforcing stimulus, and the number of trials and or latency to reach the reward is measured as outcomes. One potential finding in a behavioral toxicology study could be that subjects exposed to the chemical (i.e., treated group) required a greater number of trials and/or a longer latency to learn to turn to the correct side of the maze (i.e., the side in which the reinforcer is placed). While this might be interpreted as a learning deficit, numerous different behavioral mechanisms could actually be influencing the outcome that would also need to be considered. For example, the chemical exposure itself may impair motor function such that it takes treated subjects a longer time to traverse the maze and thus delays the time to reward (alters antecedent conditions) and/or increases the required effort for a reward (alters response topography). Alternatively, the treatment may impair sensory capacities of the subject, for example, olfactory or visual capabilities, thus impairing the ability of the subject to use visual or odor cues in guiding its direction (antecedent conditions). Finally, the stimulus being used to reward behavior may not in fact be reinforcing or not be sufficiently reinforcing (response consequences as a behavioral mechanism), particularly if the chemical

248

Deborah A. Cory-Slechta et al.

treatment itself induces any sickness effects. All of these impairments would lead to the outcome of greater number of trials or longer latency relative to non-treated controls, but they would not be consistent with a learning impairment. The water maze, often used for testing rodents, serves as another example. In this maze, the subject is placed in a circular pool of opaque water in which an escape platform is submerged beneath the surface; finding the escape platform serves as a negative reinforcer in providing escape from water, a non-preferred environment for rodents. Here too, motor and sensory impairments, among other factors, can influence outcome. Swimming is a highly effortful response, and thus any motor impairment produced by treatment could influence the ability to sustain a swimming response, i.e., reduce motor endurance. It has become accepted that a measure of swimming speed per se is a control for this problem; however, this does not measure changes in physical endurance which is actually requisite to the correct response. Similarly, impaired sensory capacity will alter ability of the subject to utilize environmental cues to learn the placement of the submerged platform. Olfactory cues left by rodents in these test environments have also been shown to influence outcome [17, 18]. Another physiological factor at play in water maze paradigms is hypothermia [19]. Studies have reported this effect to underlie the putative age-related learning deficits in rats reported in water maze performance, as preventing hypothermia by warming 23 months old rats between trials in a water maze task was shown to significantly improve their performance levels to that of young animals. As noted by the authors, these results suggest that age-related deficits in the water maze during aging are not due to the loss of visual acuity, which did not influence performance; but, as a specific measure of cognitive function, performance in the water maze can be confounded by the loss of thermoregulatory control [20]. Furthermore, water maze utilizes an escape response, i.e., negative reinforcement, which studies suggest may result in differential behavioral effects than occurs under conditions of positive reinforcement [21]. More complex behavioral techniques, albeit more expensive and time-consuming and requiring more behavioral expertise, have the capacity to address questions about specificity of the effect observed, and to concomitantly rule out potential confounding behavioral mechanisms. One such example for learning is the multiple schedule of repeated learning and performance. Initially developed for human subjects by Boren, it has since been widely used to study the effects of drugs on learning [22, 23] and to be useful across numerous species including human, monkeys, rats, and mice [24–31]; performance on this behavioral baseline also shows correlation with IQ levels in humans [29]. One version of this behavioral paradigm requires the organism to execute response sequences, i.e.,

Assessment of Neurotoxicant-Induced Changes in Behavior: Issues Related. . .

249

a designated chain of responses, for reward. In the repeated learning component (i.e., the RL component) of the session, the correct sequence changes at each successive experimental session. The RL component alternates with another component during the session, the performance (P) component. In contrast to the RL component, the response sequence reinforced during the performance component remains constant across sessions, so the organism is simply performing an already learned or rote response sequence. By alternating the repeated learning and performance components within each experimental session, e.g., after every tenth reinforcer delivery, with component changes signaled by a change in, e.g., illumination or an auditory cue, drug- or toxicantinduced changes in learning can be differentiated from nonspecific changes in motor function or motivation, i.e., confounding alternative behavioral mechanisms can be considered and/or eliminated. This is based on the premise that if a compound selectively affects learning, then decreases in accuracy should only be seen during the repeated learning components of the schedule, as no learning at all is required in the performance components. However, if decrements in accuracy also occur in the performance components, it would suggest that nonspecific influences are occurring, as intact motor, sensory, and motivational functions are required in both the repeated learning and performance components in the execution of the response and impairments of these functions would result in decreases in accuracy in both the repeated learning and performance components. Our laboratory used this paradigm to examine whether effects of developmental lead exposure on learning represented selective learning impairments [32]. Figure 5 compares cumulative records of the behavior of a typical control rat (top) to that of a rat exposed to 250 ppm lead acetate from weaning. The session began with a performance (P1) component, followed by the repeated learning component (RL1), a return to the performance component (P2), and a final presentation of the repeated learning component (RL2); components alternated after 25 reinforcer deliveries and were accompanied by a change in the lights illuminated within the chamber. Reinforcement followed each completion of a correct three-response sequence. Following extensive training on the schedule, the control rat (top record) had high levels of accuracy in P1, generating a high rate of reinforcement delivery and a relatively low rate of errors, as expected. The onset of RL1, signaled to the subject by a change in illumination within the operant chamber, required the rat to now learn a new three-response sequence that differed from the correct RL sequence from the prior session. As expected, the error rate was initially high, producing a lower rate of reinforcement delivery, but by the end of RL1, the rate of errors had begun to decline, and the rat was earning food deliveries at a faster rate as it gradually began to learn the correct

250

Deborah A. Cory-Slechta et al.

Fig. 5 Cumulative records of performance of a control rat (top) and a rat exposed from weaning to 250 ppm lead acetate in drinking water (bottom). Correct responses cumulate vertically, while errors are depicted on the bottom line of each record. RL Repeated learning; P Performance

sequence. The learning process was even more pronounced during RL2, as the rat continued to learn the correct RL sequence for this particular session. The bottom record shows a dramatic impact of lead exposure that was selective for the RL components of the schedule. Specifically, the lead-exposed rat earned virtually no food deliveries during either the RL1 or RL2 components, despite emitting hundreds of incorrect responses over the course of these components; in contrast, it exhibited high accuracy levels during both the P1 and P2 components, rapidly earning all 25 available reinforcers and eliminating potential confounding mechanisms as the explanation for the deficits in the RL components. In the case of methylmercury exposure of mice, behavioral changes in the repeated acquisition procedure were found instead to be due to motoric effects which was only possible to evaluate with the inclusion of a performance component [33]. Thus, without experimental control over behavior, such as in uncomplicated behavioral procedures, it is not possible to rule out these other potential determinants. Such alternative explanations for ostensible learning changes can apply to other negative reinforcement-based techniques as well, potentially confounding interpretations of behavioral outcome and/or effects of the treatment of interest. For example, studies have demonstrated significant increases in peripheral corticosterone levels during performance of a shuttle-box avoidance task [34– 36]. In another study in Wistar rats, swimming in a water maze caused a significantly increased release of other hormones, namely, vasopressin (AVP) within the paraventricular nucleus of the

Assessment of Neurotoxicant-Induced Changes in Behavior: Issues Related. . .

251

hypothalamus and of oxytocin (OXT) within the supraoptic nuclei on each of the three test sessions. Furthermore, plasma ACTH and corticosterone were found to be similarly elevated in response to water maze testing during each of the test sessions [37]. In another study [35], corticosterone concentrations increased when extinction of a previously rewarded response was imposed, but not when positive reinforcement continued. Activation of stress or endocrine systems modify the physiological environment and thus the context under which learning is being assayed, which can have significant consequences if these changes vary by treatment, i.e., interact with the neurotoxicant, and/or further differ by sex. If, for example, the corticosterone release is requisite to learning of the avoidance task, but is reduced by the neurotoxicant, then the framework for learning has been altered. In relation to sex differences, a study in Long-Evans rats found that corticosterone was highly elevated during and after water maze training, with females showing greater increases than males, and strong inverse correlations were observed between corticosterone and measures of water maze performance in females [38]. While such negative reinforcement-based fear conditioning behavior is typically acquired more rapidly than positive reinforcement-based behaviors, it is important to remember that the former are not in fact models of voluntary behavior [21]. Fear conditioning has been a widely used technique based on aversive stimulus presentation to study learning/memory. Like water maze or forced swimming procedures, fear conditioning also evokes significant corticosterone release. For example, a study comparing the role of shock intensity in fear conditioning found that fear conditioning resulted in corticosterone release regardless of the shock intensity used [39]. Similarly, foot-shocked rats showed higher corticosterone levels than controls that were not footshock conditioned [40]. Negative reinforcement-based shock avoidance techniques have similar consequences [41]. An additional factor that could influence learning/memory in shockbased techniques that have long been known to depend upon shock intensity [42], is sensitivity to pain; should the chemical treatment alter pain sensitivity, it would decrease learning in this paradigm. Given this, it becomes necessary to measure toxicantinduced changes in baseline shock sensitivity which is almost invariably never considered. Thus, testing in a negative reinforcement paradigm where the reinforcer consists of removal of an aversive stimulus, or shock presentation paradigms, can be associated not only with a significant activation of the hypothalamic–pituitary–adrenal axis but also with an intrahypothalamic release of AVP and OXT. Such increases in stress physiology have the potential to interact with the independent variable, particularly in the case of a chemical exposure. This could mean that the outcomes of the study do not reflect the effects of the chemical exposure per se, but the interaction of the treatment

252

Deborah A. Cory-Slechta et al.

and stress hormones if one is examining effects in behaviorally tested subjects. In summary, any behavioral assessment requires that other behavioral mechanisms of effect need to be considered. This may be particularly true for negative reinforcement and shock presentation-based behavioral paradigms and for simpler behavioral assays. But even with positive reinforcement, a stimulus usually rewarding may not increase the frequency of a response in a subject that is physically ill from a chemical exposure. In fact, if the subject is sick, lethargic, having seizures, etc. this precludes any cognitive, motor, or social testing. Ruling out alternative explanations is critical for interpretation of the behavioral changes observed. It should also be noted that many of the uncomplicated behavioral procedures are one-time snapshots taken without providing the sort of behavioral history that can reduce variability and the influence of other extraneous variables like the handler, time of day, last feeding, etc.

5

Behavioral History and Behavioral Testing Effects It is frequently the case in studies of behavioral toxicology that animals from a given exposure cohort are subjected to multiple behavioral tests within an experiment. This approach is often adopted as it results in an efficiency as it relates to costs of such experiments, as well as, to the time required. However, as has been described in numerous prior reports, experience of an organism in a given behavioral test paradigm can have significant “carry-over” effects and influence performance not only in subsequent behavioral tests, but can even alter the dose-effect curves for drug effects on subsequent behavioral tests. For example, a comparison of performances on various behavioral procedures in mice tested on a battery of tests as compared to use of mice that were naı¨ve for each of the behavioral procedures used, showed multiple differences, particularly for behavioral paradigms such as open field, rotarod, and the hot plate test [43]. This study likewise found order effects when the sequence of behavioral tests was rearranged. In another such study assessing affective behavior in rats, performance in a forced swim test was found to be dependent upon the order in which it occurred in the test battery. Furthermore, repeated testing in the open field or in the forced swim paradigm resulted in significant behavioral changes relative to earlier performances in these paradigms [44] showing the influence of repeated testing. Such changes are quite apparent in comparison to video recordings of behavior in such tests across sessions; it underscores the importance of watching what behavior is occurring in a test. A study comparing performance of C57Bl/6J mice that were experienced vs. comparable aged mice that were

Assessment of Neurotoxicant-Induced Changes in Behavior: Issues Related. . .

253

naı¨ve to a test battery comprised of the elevated plus maze test, the dark-light test, an open field test, and a novel cage test found that prior experience with the battery significantly reduced exploratory behavior and open arm-related measures. Moreover, these effects of prior experience were significantly more pronounced in older (13 weeks) as compared to younger (9 week) mice [45]. Such behavioral experience can then interact with the chemical exposure. In support of such possibilities, we found evidence for effects of prior behavioral experiences on subsequent complex learning. Rats were given the experience of either a single restraint stress followed by a single forced swim test (R + FS), or the behavioral experience of receiving food reward for lever pressing on a fixed interval (FI) schedule of reinforcement. Learning was then subsequently assessed in all groups using a repeated learning paradigm that required acquisition of a new two-response chain in each successive behavioral test session. Figure 6 depicts three outcome measures across six sessions for these offspring. As they show, accuracy tended to be lower in groups with R + FS experience than with FI experience. In addition, under those conditions, marked reductions in percent correct learning were found in both males and females subjected to the forced swim test in the two-response chain repeated learning paradigm, particularly in the early test sessions, which was accompanied by marked increases in the first response of the sequence being incorrect. Thus, an early powerful stressor had a greater impact than did early experience with positively reinforced behavior in terms of subsequent acquisition of a behavioral chain. Prior behavioral experience can also markedly alter the dose response curves for drug effects in subsequent behavioral tests. For example, a comparison of the effects of d- amphetamine on punished responding was measured in squirrel monkeys in which food-maintained responding was suppressed by the presentation of electric shock (punishment). Two monkeys were experimentally naive and two had a prior history of responding under conditions of both shock postponement (avoidance behavior) and shock presentation (punishment) schedules. D-amphetamine did not increase punished responding by naive monkeys, but markedly increased rates in the behaviorally experienced group [46]. Other examples include studies reporting that prior experience in a light/dark exploration test resulted in chlordiazepoxide-induced increases in exploration in the light compartment in naı¨ve subjects, whereas the drug had no effects in test-experienced mice [47]. A recent study found that testing in the light, as compared to the dark phase, increased locomotor activity as well as the response to d-amphetamine, while reducing social approach behavior [48]. Such findings collectively suggest that in a study design in which multiple behavioral tests are administered, the effects of a toxicant upon behavior may well depend upon the specific

254

Deborah A. Cory-Slechta et al.

Fig. 6 Changes in overall accuracy (left column), overall response rates (middle column and percent of first responses in a two-response behavioral chain that were incorrect (right column) in females (top row) and male (bottom row) rats that had prior experience of either food reward on a fixed interval schedule of reinforcement (FI) or a single exposure to restraint stress followed by forced swim (R + FS). The correct two-response changed with every session. * ¼ significant difference between FI and R + FS groups at the indicated session. Cory-Slechta et al. unpublished data

behavioral paradigms that are used, as well as the order in which the testing occurs. As a further problem, behavioral history can interact with toxicant treatment. For example, consider an experimental design in which pain is tested on a hot plate followed by a fear conditioning procedure. Assume the toxicant modifies pain sensitivity and thus hot plate performance. This modified behavioral history is now carried forward into the fear conditioning paradigm. If differences between toxicant-treated and control subjects are found in the fear conditioning paradigm, it may not be possible to ascertain whether these reflect the effect of the toxicant itself, the altered behavioral history, or perhaps an interaction of the two. Such effects may be particularly pronounced with fear conditioning. Prior studies in rats

Assessment of Neurotoxicant-Induced Changes in Behavior: Issues Related. . .

255

have shown that even a single footshock can produce a persistent hypoactivity in exposure to subsequent unknown environments [49] which in a learning test could slow the time to reward. For such reasons, it is equally important to recognize that the trajectory of behavioral responses and behavioral experience is dynamic and life-long.

6

Behavioral Testing Itself Modifies the Brain A frequent experimental design in behavioral neurotoxicology involves chemical exposure followed by behavioral testing and subsequent collection of brains from behaviorally tested subjects to be examined, for example, for potential mechanisms of the observed behavioral deficits or perhaps an intervention to mitigate the adverse behavioral effects. This approach is typical, as it may be considered both more efficient and economical in terms of animal costs. Unfortunately, this approach is confounded by the fact that, as has long been known but subsequently forgotten, behavioral experience itself modifies the brain [50–52], which would mean that changes in the brain of organisms with a behavioral history reflect the interaction of the chemical exposure with the behavioral experience, rather than simply being the result of the chemical exposure alone. The plasticity of the brain in response to behavioral experience was demonstrated more than 50 years ago. For example, early studies in this area showed systematic differences in cortical acetylcholinesterase enzyme activity not only in rats tested in different behavioral apparati, but also in comparison to rats not subjected to behavioral testing at all [53]. Furthermore, increases in rates of protein synthesis and protein levels and in amounts of RNA were found in rats from an enriched environment, and maze training itself increased cortical RNA:DNA ratios and such differences were subsequently shown to be induced across ages, although slower to emerge in later ages. Moreover, these changes can occur rapidly, being reported after only 4 days of differential housing in one study [54] and after only 40 min of experience in another report [55]. Such plasticity has been reported across mammalian species, as well as in birds, fish, fruit flies, and spiders [56]. Subsequent control experiments have confirmed that these effects did not appear to be attributable to differences in handling, stress, accelerated maturation, differential locomotion, or hormonal mediation [57]. Given this potential confound, we have implemented the inclusion of a non-behavioral control group in our experiments, as examination of brains in this group following chemical treatments should reflect actual potential sites and mechanisms of its toxicity with both groups having equivalent housing conditions. Moreover,

256

Deborah A. Cory-Slechta et al.

given that we have also found behavioral experience to mitigate the effects of exposures in our studies of neurodevelopmental toxicity of lead [58], this approach permits assessment of specific behavioral experiences that may accelerate such reversals. Of course, this also raises the issue of what constitutes a “non-behavioral” control group for such purposes and what experience best translates to human conditions. In further support of such differences, we have found evidence for effects of prior behavioral experiences on subsequent neurotransmitter changes that further interacted with prior developmental exposures to lead and/or prenatal stress. We exposed mice to 0 or 100 ppm lead (0 vs. 100) in drinking water from 2 to 3 mos prior to breeding through lactation, with half of the dams in each lead group also receiving prenatal stress (NS ¼ no stress; PS ¼ prenatal stress). Subsequently, offspring from each of these groups were given either a series of four forced swim (FS) experiences, or the behavioral experience of receiving food reward for lever pressing on a fixed interval (FI) schedule of reinforcement, leading to a total of eight groups of male and of female offspring (0-NS-FI, 0-NS-FS, 0-PS-FI, 0-PS-FS, 100-NS-FI, 100-NS-FS, 100-PS-FI, 100-PS-FS) [59]. Figure 7 depicts changes in brain dopamine concentrations in four different brain regions of male mice following developmental lead exposure and/or prenatal stress and subsequent FI vs. FS experience. Here, behavioral history interacted with developmental lead and/or prenatal stress to markedly modify brain dopamine relative to 0-NS control values. This included marked increases in frontal cortex brain dopamine after FI testing following lead exposure, but particularly in organisms that had been exposed to both lead and prenatal stress, whereas increases in dopamine concentrations were seen in response to FS experience in groups receiving either prenatal stress only or Pb only in striatum, midbrain, and hippocampus. In another such study from our laboratory in which rats were maternally exposed to lead (0, 50, or 150 ppm in drinking water) and/or prenatal stress, offspring were given experience on the FI schedule of food reward, as previously described, or no behavioral experience (NFI) and subsequent brain neurochemical alterations were measured [58]. As can be seen in Fig. 8, a similar interaction of behavioral history occurred with developmental insults, with effects in both sexes. For example, increases in nucleus accumbens norepinephrine concentrations occurred in females with developmental exposures to lead and to lead + prenatal stress only if they also had an FI behavioral history. Frontal cortex dopamine turnover concentrations were markedly increased only in mice whose dams received 50 ppm lead plus prenatal stress, or the 150 ppm exposure level and/or prenatal stress followed by no behavioral experience. In striatum, dopamine concentrations of males with prior FI experience only were elevated if they had maternal exposures to 50 ppm

Assessment of Neurotoxicant-Induced Changes in Behavior: Issues Related. . .

257

Fig. 7 Dopamine concentrations in indicated brain regions of male mice exposed to maternal lead exposure (0 or 100 ppm in drinking water and/or prenatal stress followed by behavioral experience on either a fixed interval schedule of food reward (FI) or a series of four forced swim experiences (FS). Modified from CorySlechta et al. (2013). * ¼ significantly different from corresponding treatments in the other behavioral experience condition

lead plus prenatal stress, or the 150 ppm exposure level and/or prenatal stress, whereas dopamine turnover was markedly increased by developmental lead and/or prenatal stress in males that also had no behavioral experience. Unintended behavioral histories may also be produced by variations in housing conditions for laboratory animals, particularly rodents. A recent emphasis over the past decade has been the implementation of “enriched” housing as a mechanism to improve animal welfare within vivarium housing. Such enrichment has ranged from marked differences in cage sizes and numbers of rodents within a cage, to provisions of nesting materials, different bedding, and various physical structures. However, such “enrichment” may have very unintended and unrecognized consequences for studies of behavioral toxicity, as such types of enrichment can produce changes in both physiology and behavior [60, 61], and as noted above, such effects may interact with the neurotoxicant being

258

Deborah A. Cory-Slechta et al.

Fig. 8 Concentrations of indicated neurotransmitters in brain regions and by sex of rats that were maternally exposed to lead in drinking water (0, 50, or 150 ppm) and/or prenatal stress followed by experience on the FI schedule or with no FI experience (NFI). * ¼ significantly different from corresponding treatments in the other behavioral experience condition. Modified from CorySlechta et al. (2009)

studied. Such interactive effects may also significantly differ by sex [62]. The use of “enriched housing” has also been reported to increase variability among animals in outcome measures [60]. Here too such differences can significantly contribute to differences across studies in reliability and reproducibility [63].

7

Use of the Same Behavioral Test Paradigm across Species One of the goals of behavioral toxicology is to ascertain the extent to which chemical exposures may impact the human brain and behavior, i.e., a translational objective. For such reasons, an approach which could significantly enhance the translational nature

Assessment of Neurotoxicant-Induced Changes in Behavior: Issues Related. . .

259

of behavioral toxicology experiments derived from animal models is the use of the same behavioral paradigms across species. Such an approach incorporates both forward translation and backward translational capabilities into our understanding. For example, findings in experimental animal studies can be forward translated into subsequent epidemiological studies in populations exposed to the chemical of interest. Furthermore, information from human behavioral testing can be pursued in animal models to elaborate mechanisms of effect, biomarkers, and potential behavioral intervention or therapeutic strategies. Importantly, the use of the same paradigm, rather than different tests stated to measure the same domain, eliminates many of the questions that arise about extrapolation across different behavioral paradigms. One such example was the use of repeated acquisition paradigm across species cited above [22–31]. A wide of cross-species behavioral procedures are available and have been described in prior reports [64–66].

8

Summary and Conclusions In conclusion, studies that report alterations in learning following toxicant exposure, drug administration, genetic modification, etc. need to consider that behavior is the product of an organism’s physiology and environment. Further, the equation is not static, as developmental and behavioral histories feedback on one another across the life span. Context matters in discussions of predictability and reproducibility of behavioral results. Given that behavioral experience can modify the effects of treatment, researchers must describe behavioral histories for appropriate comparisons. Additionally, the effects of a treatment on the brain following behavioral experience represent the effects of both the treatment and behavioral experience. Furthermore, understanding the behavioral mechanisms of learning is critical in ensuring that confounds, such as alterations in motivation (deprivation/satiation), motor function, pain sensitivity, and sex differences in stress responsivity, are controlled for to rule out these effects as primary contributors to toxicity, and not learning. In these cases, there is no substitute for a performance component or additional behavioral assays that ensure other behavioral domains are not altered. To this end, there is also no substitute for watching your test subjects. Understanding the evolutionary history of the species being utilized will help improve learning assays by increasing reward frequencies through more naturalistic behavioral responses such as nose-pokes or jumping. With these considerations included in experimental designs, behavior, and learning paradigms specifically remain a critical and strong endpoint for translation across species for neurotoxicology.

260

Deborah A. Cory-Slechta et al.

References 1. Dick DM (2011) Gene-environment interaction in psychological traits and disorders. Annu Rev Clin Psychol 7:383–409. https://doi.org/ 10.1146/annurev-clinpsy-032210-104518 2. Wingfield JC, Hegner RE, Dufty AM, Ball GF (1990) The “challenge hypothesis”: theoretical implications for patterns of testosterone secretion, mating systems, and breeding strategies. Am Nat 136(6):829–846 3. Bateson P, Laland KN (2013) Tinbergen’s four questions: an appreciation and an update. Trends Ecol Evol 28(12):712–718. https:// doi.org/10.1016/j.tree.2013.09.013 4. Skinner BF (1953) Science and human behavior. Simon & Schuster Inc., New York 5. Buccafusco JJ (2009) Methods of behavior analysis in neuroscience, 2nd edn. CRC Press/Taylor & Francis, Boca Raton, FL 6. Nestler EJ, Hyman SE (2010) Animal models of neuropsychiatric disorders. Nat Neurosci 13 (10):1161–1169. https://doi.org/10.1038/ nn.2647 7. Tinbergen N (1963) On aims and methods of ethology. Z Tierpsychol 20(4):410–433. https://doi.org/10.1111/j.1439-0310.1963. tb01161.x 8. Kenward B, Rutz C, Weir AAS, Kacelnik A (2006) Development of tool use in new Caledonian crows: inherited action patterns and social influences. Anim Behav 72 (6):1329–1343. https://doi.org/10.1016/j. anbehav.2006.04.007 9. Laties VG (1975) The role of discriminative stimuli in modulating drug action. Fed Proc 34(9):1880–1888 10. Wood RW, Rees DC, Laties VG (1983) Behavioral effects of toluene are modulated by stimulus control. Toxicol Appl Pharmacol 68 (3):462–472. https://doi.org/10.1016/ 0041-008x(83)90291-0 11. Gerhardt S, Liebman JM (1981) Differential effects of drug treatments on nose-poke and bar-press self-stimulation. Pharmacol Biochem Behav 15(5):767–771. https://doi.org/10. 1016/0091-3057(81)90020-4 12. Riess D (1971) Shuttleboxes, Skinner boxes, and Sidman avoidance in rats: acquisition and terminal performance as a function of response topography. Psychon Sci 25(5):283–286. https://doi.org/10.3758/BF03335878 13. Barrett JE, Valentine JO, Katz JL (1981) Effects of chlordiazepoxide and d-amphetamine on responding of squirrel monkeys maintained under concurrent or second-order schedules of response-produced

food or electric shock presentation. J Pharmacol Exp Ther 219(1):199 14. Lett BT (1975) Long delay learning in the T-maze. Learn Motiv 6(1):80–90. https:// doi.org/10.1016/0023-9690(75)90036-3 15. Messing RB, Kleven MS, Sparber SB (1986) Delaying reinforcement in an autoshaping task generates adjunctive and superstitious behaviors. Behav Process 13(4):327–338. https://doi.org/10.1016/0376-6357(86) 90028-8 16. Saltzman IJ (1951) Delay of reward and human verbal learning. J Exp Psychol 41(6):437–439 17. Roullet P, Lassalle JM, Jegat R (1993) A study of behavioral and sensorial bases of radial maze learning in mice. Behav Neural Biol 59 (3):173–179. https://doi.org/10.1016/ 0163-1047(93)90926-9 18. Means LW, Alexander SR, O’Neal MF (1992) Those cheating rats: male and female rats use odor trails in a water-escape “working memory” task. Behav Neural Biol 58(2):144–151 19. Panakhova E, Buresova O, Bures J (1984) The effect of hypothermia on the rat’s spatial memory in the water tank task. Behav Neural Biol 42(2):191–196. https://doi.org/10.1016/ s0163-1047(84)91059-8 20. Lindner MD, Gribkoff VK (1991) Relationship between performance in the Morris water task, visual acuity, and thermoregulatory function in aged F-344 rats. Behav Brain Res 45 (1):45–55 21. Magoon MA, Critchfield TS, Merrill D, Newland MC, Schneider WJ (2017) Are positive and negative reinforcement “different”? Insights from a free-operant differential outcomes effect. J Exp Anal Behav 107 (1):39–64. https://doi.org/10.1002/jeab. 243 22. Boren JJ (1963) Repeated acquisition of new behavioral chains. Am Psychol 17:421 23. Boren JJ, Devine DD (1968) The repeated acquisition of behavioral chains. J Exp Anal Behav 11:651–660 24. Brooks AI, Cory-Slechta DA, Murg SL, Federoff HJ (2000) Repeated acquisition and performance chamber for mice: a paradigm for assessment of spatial learning and memory. Neurobiol Learn Mem 74(3):241–258 25. Brooks AI, Stein CS, Hughes SM, Heth J, McCray PM Jr, Sauter SL, Johnston JC, Cory-Slechta DA, Federoff HJ, Davidson BL (2002) Functional correction of established central nervous system deficits in an animal model of lysosomal storage disease with feline

Assessment of Neurotoxicant-Induced Changes in Behavior: Issues Related. . . immunodeficiency virus-based vectors. Proc Natl Acad Sci U S A 99(9):6216–6221 26. Arnold MA, Newland MC (2018) Variable behavior and repeated learning in two mouse strains: developmental and genetic contributions. Behav Process 157:509–518. https:// doi.org/10.1016/j.beproc.2018.06.007 27. Johnson JM, Bailey JM, Johnson JE, Newland MC (2010) Performance of BALB/c and C57BL/6 mice under an incremental repeated acquisition of behavioral chains procedure. Behav Process 84(3):705–714. https://doi. org/10.1016/j.beproc.2010.04.008 28. Shen AN, Pope DA, Hutsell BA, Newland MC (2015) Spatial discrimination reversal and incremental repeated acquisition in adolescent and adult BALB/c mice. Behav Process 118:59–70. https://doi.org/10.1016/j. beproc.2015.06.005 29. Baldwin RL, Chelonis JJ, Prunty PK, Paule MG (2012) The use of an incremental repeated acquisition task to assess learning in children. Behav Process 91(1):103–114. https://doi. org/10.1016/j.beproc.2012.06.004 30. Ferguson SA, Gopee NV, Paule MG, Howard PC (2009) Female mini-pig performance of temporal response differentiation, incremental repeated acquisition, and progressive ratio operant tasks. Behav Process 80(1):28–34. https://doi.org/10.1016/j.beproc.2008.08. 006 31. Garey J, Paule MG (2010) Effects of chronic oral acrylamide exposure on incremental repeated acquisition (learning) task performance in Fischer 344 rats. Neurotoxicol Teratol 32(2):220–225. https://doi.org/10. 1016/j.ntt.2009.10.001 32. Cohn J, Cox C, Cory-Slechta DA (1993) The effects of lead exposure on learning in a multiple repeated acquisition and performance schedule. Neurotoxicology 14:329–346 33. Bailey JM, Hutsell BA, Newland MC (2013) Dietary nimodipine delays the onset of methylmercury neurotoxicity in mice. Neurotoxicology 37:108–117. https://doi.org/10.1016/j. neuro.2013.03.011 34. Coover GD, Hart RP, Frey MJ (1986) Corticosterone, free fatty acid and glucose responses of rats to footshock, fear, novel stimuli and instrumental reinforcement. Psychoneuroendocrinology 11(3):373–388 35. de Boer SF, de Beun R, Slangen JL, van der Gugten J (1990) Dynamics of plasma catecholamine and corticosterone concentrations during reinforced and extinguished operant behavior in rats. Physiol Behav 47

261

(4):691–698. https://doi.org/10.1016/ 0031-9384(90)90079-J 36. Port RL, Sisak ME, Finamore TL, Soltrick ML, Seybold KS (1998) Role of corticosterone in extinction of an appetitive instrumental response. Int J Neurosci 96(1–2):13–21. https://doi.org/10.3109/ 00207459808986454 37. Engelmann M, Ebner K, Landgraf R, Wotjak CT (2006) Effects of Morris water maze testing on the neuroendocrine stress response and intrahypothalamic release of vasopressin and oxytocin in the rat. Horm Behav 50 (3):496–501. https://doi.org/10.1016/j. yhbeh.2006.04.009 38. Beiko J, Lander R, Hampson E, Boon F, Cain DP (2004) Contribution of sex differences in the acute stress response to sex differences in water maze performance in the rat. Behav Brain Res 151(1):239–253. https://doi.org/10. 1016/j.bbr.2003.08.019 39. Pietersen CY, Bosker FJ, Postema F, den Boer JA (2006) Fear conditioning and shock intensity: the choice between minimizing the stress induced and reducing the number of animals used. Lab Anim 40(2):180–185. https://doi. org/10.1258/002367706776319006 40. Finn DP, Jhaveri MD, Beckett SR, Madjd A, Kendall DA, Marsden CA, Chapman V (2006) Behavioral, central monoaminergic and hypothalamo-pituitary-adrenal axis correlates of fear-conditioned analgesia in rats. Neuroscience 138(4):1309–1317. https://doi.org/10. 1016/j.neuroscience.2005.11.063 41. Wade S (1984) Corticosterone availability in male rats at rest and during shock avoidance: differences among individuals, temporal patterns and effects of dexamethasone. J Endocrinol 103(2):187–194. https://doi.org/10. 1677/joe.0.1030187 42. Davis M, Astrachan DI (1978) Conditioned fear and startle magnitude: effects of different footshock or backshock intensities used in training. J Exp Psychol Anim Behav Process 4 (2):95–103. https://doi.org/10.1037/00977403.4.2.95 43. McIlwain KL, Merriweather MY, Yuva-Paylor LA, Paylor R (2001) The use of behavioral test batteries: effects of training history. Physiol Behav 73(5):705–717 44. Blokland A, Ten Oever S, van Gorp D, van Draanen M, Schmidt T, Nguyen E, Krugliak A, Napoletano A, Keuter S, Klinkenberg I (2012) The use of a test battery assessing affective behavior in rats: order effects. Behav Brain Res 228(1):16–21. https://doi.org/10. 1016/j.bbr.2011.11.042

262

Deborah A. Cory-Slechta et al.

45. von Kortzfleisch VT, Kastner N, Prange L, Kaiser S, Sachser N, Richter SH (2019) Have I been here before? Complex interactions of age and test experience modulate the results of behavioural tests. Behav Brain Res 367:143–148. https://doi.org/10.1016/j. bbr.2019.03.042 46. Barrett JE (1977) Behavioral history as a determinant of the effects of d-amphetamine on punished behavior. Science 198(4312):67. https://doi.org/10.1126/science.408925 47. Holmes A, Iles JP, Mayell SJ, Rodgers RJ (2001) Prior test experience compromises the anxiolytic efficacy of chlordiazepoxide in the mouse light/dark exploration test. Behav Brain Res 122(2):159–167 48. Richetto J, Polesel M, Weber-Stadlbauer U (2019) Effects of light and dark phase testing on the investigation of behavioural paradigms in mice: relevance for behavioural neuroscience. Pharmacol Biochem Behav 178:19–29. https://doi.org/10.1016/j.pbb.2018.05.011 49. Daviu N, Fuentes S, Nadal R, Armario A (2010) A single footshock causes long-lasting hypoactivity in unknown environments that is dependent on the development of contextual fear conditioning. Neurobiol Learn Mem 94 (2):183–190. https://doi.org/10.1016/j. nlm.2010.05.005 50. Rosenzweig MR (2007) Frontiers in neuroscience. Modification of brain circuits through experience. In: Bermudez-Rattoni F (ed) Neural plasticity and memory: from genes to brain imaging. CRC Press/Taylor & Francis Group, LLC., Boca Raton, FL 51. Bennett EL, Diamond MC, Krech D, Rosenzweig MR (1964) Chemical and anatomical plasticity of brain. Science 146(3644):610. https://doi.org/10.1126/science.146.3644. 610 52. Kozorovitskiy Y, Gross CG, Kopil C, Battaglia L, McBreen M, Stranahan AM, Gould E (2005) Experience induces structural and biochemical changes in the adult primate brain. Proc Natl Acad Sci U S A 102 (48):17,478–17,482. https://doi.org/10. 1073/pnas.0508817102 53. Rosenzweig MR, Krech D, Bennett EL, Diamond MC (1962) Effects of environmental complexity and training on brain chemistry and anatomy: a replication and extension. J Comp Physiol Psychol 55(4):429–437. https://doi.org/10.1037/h0041137 54. Wallace CS, Kilman VL, Withers GS, Greenough WT (1992) Increases in dendritic length in occipital cortex after 4 days of differential housing in weanling rats. Behav Neural Biol

58(1):64–68. https://doi.org/10.1016/ 0163-1047(92)90937-y 55. Ferchmin PA, Eterovic VA (1986) Forty minutes of experience increase the weight and RNA content of cerebral cortex in periadolescent rats. Dev Psychobiol 19(6):511–519. https://doi.org/10.1002/dev.420190604 56. Mohammed AH, Zhu SW, Darmopil S, Hjerling-Leffler J, Ernfors P, Winblad B, Diamond MC, Eriksson PS, Bogdanovic N (2002) Environmental enrichment and the brain. Prog Brain Res 138:109–133. https://doi.org/10. 1016/s0079-6123(02)38074-9 57. Lim R (1970) Grant Newton and Seymour Levine (Eds.). Early experience and behavior (the psychobiology of development). Springfield, III: Charles C Thomas, 1968. Behav Sci 15(4):367–369. https://doi.org/10.1002/bs. 3830150411 58. Cory-Slechta DA, Virgolini MB, RossiGeorge A, Weston D, Thiruchelvam M (2009) Experimental manipulations blunt time-induced changes in brain monoamine levels and completely reverse stress, but not Pb+/stress-related modifications to these trajectories. Behav Brain Res 205(1):76–87 59. Cory-Slechta DA, Merchant-Borna K, Allen J, Liu S, Weston D, Conrad K (2012) Variations in the nature of behavioral experience can differentially alter the consequences of developmental exposures to Lead. Prenatal Stress Combination Toxicol Sci. https://doi.org/ 10.1093/toxsci/kfs260 60. Benefiel AC, Dong WK, Greenough WT (2005) Mandatory “enriched” housing of laboratory animals: the need for evidence-based evaluation. ILAR J 46(2):95–105. https:// doi.org/10.1093/ilar.46.2.95 61. Bayne K (2005) Potential for unintended consequences of environmental enrichment for laboratory animals and research results. ILAR J 46(2):129–139. https://doi.org/10.1093/ ilar.46.2.129 62. Juraska JM (1998) Neural plasticity and the development of sex differences. Annu Rev Sex Res 9:20–38 63. Crabbe JC, Wahlsten D, Dudek BC (1999) Genetics of mouse behavior: interactions with laboratory environment. Science 284 (5420):1670–1672. https://doi.org/10. 1126/science.284.5420.1670 64. Wallace TL, Ballard TM, Glavis-Bloom C (2015) Animal paradigms to assess cognition with translation to humans. Handb Exp Pharmacol 228:27–57. https://doi.org/10.1007/ 978-3-319-16522-6_2

Assessment of Neurotoxicant-Induced Changes in Behavior: Issues Related. . . 65. Brown VJ, Tait DS (2016) Attentional set-shifting across species. Curr Top Behav Neurosci 28:363–395. https://doi.org/10. 1007/7854_2015_5002 66. Barnett JH, Blackwell AD, Sahakian BJ, Robbins TW (2016) The paired associates learning

263

(PAL) test: 30 years of CANTAB translational neuroscience from laboratory to bedside in dementia research. Curr Top Behav Neurosci 28:449–474. https://doi.org/10.1007/ 7854_2015_5001

Part IV Molecular Methods

Chapter 12 Assessment of Neurofilament Light Protein as a Serum Biomarker in Rodent Models of Toxic-Induced Peripheral Neuropathy Giulia Fumagalli, Guido Cavaletti, Henrik Zetterberg, and Cristina Meregalli Abstract Chemotherapy-Induced Peripheral Neurotoxicity (CIPN) is a side effect frequently caused by common antitumor drugs, which may induce a severe and persistent limitation of the quality of life of cancer patients. Today, nerve conduction studies are considered the most objective indicators for CIPN diagnosis. Unfortunately, they are not easily available at most oncology centers. Therefore, a noninvasive and highly sensitive method is required to confirm nerve damage. Increased evidence supports the potential utility of fluid-based biomarkers to predict tissue damage and to monitor neurotoxicity due to drug administration or the efficacy of disease-modifying treatments. Neurofilaments, the major intermediate filaments in neurons that are specifically expressed in axons, have been investigated as potential biomarker candidates that might be used for this purpose. Neurofilament light chain (NfL) protein is increasingly proposed as a blood biomarker in several neurological diseases mainly affecting the central nervous system. In addition, analysis of serum NfL was evaluated also in peripheral neuropathies including Guillain–Barre´ syndrome, chronic inflammatory demyelinating and vasculitic neuropathies, and Charcot–Marie–Tooth. This chapter aims to provide an overview of the methods that allow NfL quantification in serum, focusing on the most recent ultrasensitive single molecule array (Simoa) assay. This technique is likely to be the best method for NfL dosage in CIPN models to predict the onset of large caliber neuronal dysfunction. Since blood sampling is an easily accessible technique, serum NfL may provide important help to monitor neuroaxonal damage after chemotherapy treatment, and might represent promising tools to follow CIPN progress. Key words Neurofilament light, Axonal injury, Chemotherapy-induced peripheral neurotoxicity, Serum biomarker, Neurotoxicity

1

Introduction The intermediate filaments are specific neuronal proteins composing the cellular cytoskeleton, and they exert a critical role in supporting axon and dendrite outgrowth, stabilization, and function. Among intermediate filaments, neurofilaments are particularly

Jordi Llorens and Marta Barenys (eds.), Experimental Neurotoxicology Methods, Neuromethods, vol. 172, https://doi.org/10.1007/978-1-0716-1637-6_12, © Springer Science+Business Media, LLC, part of Springer Nature 2021

267

268

Giulia Fumagalli et al.

abundant and they are important for the increase of axonal caliber and consequently for the speed of electrical impulse transmission and structural stability [1, 2]. Neurofilament subunits are classified on the basis of their molecular weight as light, medium, and heavy [1]. During the last few years, several studies have indicated the quantification of serum neurofilament light chain (NfL) as an important biomarker of damage in different diseases of the central nervous system such as Alzheimer’s, Huntington’s, and Parkinson’s diseases, amyotrophic lateral sclerosis, and multiple sclerosis [2–4]. Recently, it has been demonstrated that the use of NfL concentration can be used to detect peripheral nervous system diseases, including inherited [5, 6] and acquired peripheral neuropathies [7] and vasculitic neuropathy [8]. This is due to the fact that upon axonal injury, irrespective of its cause, NfLs are released into the interstitial fluid, CSF, and blood; the magnitude of the increase correlates with the intensity and/or severity of the axonal injury process [4]. To date, grading neuroaxonal damage in a particular peripheral neurotoxicity induced by toxic agents remains an unmet clinical need. Chemotherapy-induced peripheral neurotoxicity (CIPN) is a serious side effect in patients undergoing anticancer therapies commonly used for breast, colorectal, head and neck, lung, prostate, ovarian, hematological, and testicular cancers [9]. The common anticancer drugs known to cause CIPN include platinum derivatives (e.g., cisplatin), vinka-alkaloids (e.g., vincristine), taxanes (e.g., paclitaxel), and proteasome inhibitors (e.g., bortezomib) [10]. Neurotoxicity may significantly affect patient daily activities and quality of life, leading to anticancer treatment modification or even withdrawal [11, 12]. In fact, neuroaxonal damage and loss of fibers can result in severe and/or permanent disability. For this reason, it is important to assess CIPN in a simple and reproducible manner. Currently, nerve conduction studies are the conventional method employed to objectively detect and monitor CIPN, in combination with grading scales based on subjective evaluations [13, 14]. However, this approach is not completely satisfactory in evaluating CIPN because on the one hand nerve conduction studies routinely performed deal in the detection of only large myelinated fibers whereas they are not able to detect any alteration in small fibers. In fact, techniques for assessing the involvement of small fibers are limited, and they are not employed in the standard practice [15, 16]. On the other hand, the oncological scales (e.g., the National Cancer Institute Common Toxicity Criteria, NCI CTC) usually underestimate the severity and the frequency of symptoms, and they are not always unambiguously described, leading to a variable interpretation [13]. Moreover, to allow a clinical

NfL as a Biomarker for Chemotherapy-Induced Peripheral Neurotoxicity (CIPN)

269

feasibility of an early and reliable detection of CIPN aimed at avoiding an irreversible and permanent damage, the method needs to be simple and reproducible. The use of biomarkers released into blood from injured axons would fit this aim, and additionally it would not add an invasive procedure to the standard patients’ workout as blood sampling is routinely performed in patients undergoing chemotherapy. Moreover, each subject could be examined before, during, and after chemotherapy in order to observe any change from the baseline value. In preclinical murine models of CIPN-induced by cisplatin, vincristine, and paclitaxel, we have demonstrated the possible role of serum NfL concentration as a potential biomarker for axonal damage severity [17, 18]. In particular, the observed increase in NfL concentration correlated with the severity of axonal damage demonstrated by neurophysiology and morphological investigations and with the temporal course of the pathology. Moreover, at least in some of these models, NfL levels were higher in treated animals even before overt neurophysiological and morphological changes. 1.1 Methods for NfL Detection

In the last three decades, several studies were conducted in order to develop a sensitive immunoassay for NfL detection. The methods for neurofilament protein detection could be divided in four groups, depending on assay sensitivity [4]. The first-generation immunoassays were semi-quantitative methods based on electrophoretic protein separation followed by immunoblotting. These techniques have never been employed in clinical laboratory practice due to their semi-quantitative nature and moreover they are quite laborious [19]. The sandwich enzyme-linked immunosorbent assay (ELISA) belongs to the second-generation assays. It provides quantitative data and a higher analytical accuracy of neurofilament detection. However, while this approach can quantify neurofilaments in cerebrospinal fluid, its analytical sensitivity is not good enough for blood [4, 20]. The electrochemiluminiscence (ECL) technology is a third-generation assay that is approximately 10-fold more sensitive than ELISA [21]. ECL-based measurement of NfL can quantify pathological levels of the biomarker but does not cover the normal range of values, which could be a target values in treatment studies. Our analyses were performed using an innovative technique called single molecule array (Simoa) [22]. Simoa-based measurement of NfL is basically a bead-based sandwich ELISA where the detection reaction is compartmentalized in small micro-wells (each well-fitting one single antibody-conjugated bead), which permits single molecule counting of NfL [23]. Simoa-based NfL measurement belongs to the fourth-generation NfL quantification technology, and it is 120-fold and 25-fold more sensitive than ELISA and ECL assays, respectively.

270

2

Giulia Fumagalli et al.

Materials

2.1 CIPN Animal Models

All the procedures were approved by Animal Care and Use Committee of the University of Milano-Bicocca. The experiments were performed in conformity with the institutional and governmental guidelines for human treatment of laboratory animals set forth in the Guide for the Care and Use of Laboratory Animals (Office of Laboratory Animal Welfare) as well as with the Italian D.L.vo n.26/2014 and the Europeans Union directive 2010/63/UE. Adult female Wistar rats (175–200 g, Envigo, Udine, Italy) were used in order to study CIPN. Animals were housed in a certified and limited access animal facility under constant temperature (21  C  2) and humidity (50%  20). Artificial lighting provided a 12 h light/12 h dark (7 a.m.–7 p.m.) with food and water ad libitum. Animal health condition was monitored daily and rats were sacrificed under deep anesthesia with CO2 in order to collect samples. In order to assess CIPN, three different models were obtained using repeated injections of the drugs in rats: vincristine (intravenous (i.v.) administration, 0.2 mg/kg, q7dx4 ws, TEVA Pharma B. V., Mijdrecht, The Netherlands), cisplatin (intraperitoneal (i.p.) administration, 2 mg/kg, 2qwx4 ws, Accord Healthcare Limited, Middlesex, UK) and paclitaxel (i.v. administration, 10 mg/kg, q7dx4 ws, LC laboratories, Woburn, MA, USA).

2.2 CIPN Animal Assessment

Despite the large diffusion of peripheral neurotoxicity associated with antineoplastic drugs, the molecular mechanism of the onset of this side effect remains largely unknown. Here, we report several highly reliable and reproducible rat models that effectively allow to use multimodal approaches (neurophysiological and morphological investigations) to identify the features of the peripheral neurotoxicity. In addition, our models are able to reflect multiple aspects of human peripheral sensory neurotoxicity disease since they manifested axonal degeneration, as well as neurophysiological deficits and loss of intraepidermal nerve fibers, as clinically observed. All the treatments induced CIPN of different severity depending on the chemotherapy drugs employed and their mechanism of action on peripheral nerve system. In these preclinical models, the axonal damage was evaluated by several techniques. In particular, neurophysiological analyses were performed on peripheral nerves. Briefly, conduction studies (potential amplitude and conduction velocity) of caudal nerves and hind limb digital nerves were performed through an orthodromic stimulation. During all the registration period, the animal was kept under deep isoflurane anesthesia and body temperature was maintained constant (37  C  0.5) using a heating pad. Myto II EMG apparatus (EBN Neuro, Florence, Italy) and subdermal needle

NfL as a Biomarker for Chemotherapy-Induced Peripheral Neurotoxicity (CIPN)

271

electrodes (Ambu Neuroline, Ambu, Ballerup, Denmark) were employed for all recordings [17]. The severity of axonal damage detected by neurophysiological investigations was confirmed by morphological study on collected nerves; otherwise, the evaluation of intraepidermal nerve fibers revealed a decrease of small unmyelinated fiber density as previously described [18].

3

Methods

3.1 Blood Specimen Collection and Processing

The aim of this section is to describe the procedure required to collect suitable serum sample in order to perform the Simoa assay. 1. Blood samples were collected by venipuncture 2 days after drug administration. The animals were kept in a plexiglas restrain cage and the tail was dipped in water at 37  C for few minutes in order to obtain a dilatation of the vein. 2. Blood collection was performed using a deltaven T 22G (MEDVET srl, Taranto, Italy) and then the sample was withdrawn by gravity and collected in serum-separating tubes (APTACA Spa, Canelli (AT), Italy). Almost 500 μL of blood was collected for each animal. 3. Serum was obtained through centrifugation at 4  C, 3500 g for 15 min and then aliquoted and stored at 80  C until NfL quantification.

3.2 Single-Molecule Array (Simoa) Technology

Rat serum NfL concentration was measured using an in housedeveloped Simoa assay, in which the same monoclonal antibodies (UD1 and UD2) and calibrator as in the NF-light ELISA for CSF NfL (UmanDiagnostics) were transferred onto the Simoa platform (Quanterix, Billerica, MA, USA), as previously described in detail [24]. Human and rodent NfL sequences are completely conserved and the assay works on both rat and mouse samples. 1. Paramagnetic carboxylated beads (Quanterix Corp, Boston, MA, USA) were coated with an anti-neurofilament light antibody (UD1, UmanDiagnostics, Umea˚, Sweden) and incubated 35 min with sample and a biotinylated anti-neurofilament light antibody (UD2, UmanDiagnostics) in a Simoa HD-1 instrument (Quanterix). The bead-conjugated immunocomplex was thoroughly washed before incubation with streptavidinconjugated β-galactosidase (Quanterix). 2. After additional washes, resorufin β-D-galactopyranoside (Quanterix) was added and the immunocomplex was applied to a multi-well array designed to enable imaging of every single bead to capture light emission and quantify it in digital (average number of captured enzymes per bead [AEB]) mode.

272

Giulia Fumagalli et al.

3. The AEB of samples was interpolated onto the calibrator curve constructed by AEB measurements on bovine NfL (UmanDiagnostics) serially diluted in assay diluent. 4. Samples were analyzed using one batch of reagents and animal treatment information was blinded to the one performing the analysis. The average repeatability of the assay was assessed by measurements of quality control samples and the coefficient of variation was 6.2% for a sample with a mean NfL concentration of 50.7 pg/mL, and 12.3% for a sample with a mean NfL concentration of 22.6 pg/mL. This assay set up forms the basis for the commercially available NF-Light Simoa assay (Quanterix).

4

Conclusions Over the last decades, increasing data have been supporting the use of NfL as a reliable biofluid-based biomarker for neuroaxonal injury in central nervous system diseases. NfL is a structural protein also in peripheral nerves, but the literature data regarding its potential role as a peripheral nerve injury marker is scarce and more studies on the topic are required. Conventional ELISA and ECL are the most common tests available for measuring NfL concentration in biofluids, but blood levels are too low to allow for reliable quantification, at least in the normal range [25]. In 2015, the first ultrasensitive NfL assay that allowed for reliable quantification of NfL in blood samples was published [23]. This assay is based on Simoa technology that permits single molecule counting of NfL and subfemtomolar quantification of the protein. Therefore, we used this innovative approach in order to study the onset and progression of CIPN in different preclinical models. In particular, we demonstrated the utility of Simoa for the detection and grading of axonal damage severity in a preclinical model with vincristine [17], cisplatin and paclitaxel [18], as showed in Fig. 1. Therefore, although using the same reagents as in the ELISA kit, the Simoa system provides the best analytical sensitivity across all currently available immunoassays, representing an ideal tool for implementation in clinical setting. The development of ultrasensitive blood-based detection approach will provide additional information on axon pathology by monitoring longitudinally neuronal degeneration, in order to pursue an early detection of CIPN and avoid irreversible nerve damage. Moreover, further investigations may be extended to other antineoplastic drugs and a potential use of this technique would be evaluated for monitoring the responses to therapy in trials.

NfL as a Biomarker for Chemotherapy-Induced Peripheral Neurotoxicity (CIPN)

273

Fig. 1 The graphs show the study of the axonal damage in preclinical models of neurotoxicity induced by vincristine (VCR), cisplatin (CDDP), and paclitaxel (PTX). The most significant neurophysiological results (a) and NfL dosage (b) obtained at the end of treatment are reported. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001 vs control (CTRL) (2-sided Mann–Whitney U test). Full original data of VCR model were published on Experimental Neurology [17], whereas results regarding CDDP and PTX models were published on Archives of Toxicology [18]

Acknowledgments GC is supported by the Italian PRIN grant (#2017ZFJCS3) and CM is supported by Fondazione Cariplo grant (#2019-1482) . HZ is a Wallenberg Scholar supported by grants from the Swedish Research Council (#2018-02532), the European Research Council (#681712), Swedish State Support for Clinical Research (#ALFGBG-720931) and the UK Dementia Research Institute at UCL.

274

Giulia Fumagalli et al.

Conflicts of Interest HZ has served at scientific advisory boards for Roche Diagnostics, Wave, Samumed and CogRx, has given lectures in symposia sponsored by Alzecure and Biogen, and is a co-founder of Brain Biomarker Solutions in Gothenburg AB, a GU Ventures-based platform company at the University of Gothenburg. The other authors declare that they have no conflict of interest. References 1. Perrot R, Berges R, Bocquet A et al (2008) Review of the multiple aspects of neurofilament functions, and their possible contribution to neurodegeneration. Mol Neurobiol 38 (1):27–65 2. Perrot R, Eyer J (2009) Neuronal intermediate filaments and neurodegenerative disorders. Brain Res Bull 80(4–5):282–295 3. Novakova L, Zetterberg H, Sundstro¨m P et al (2017) Monitoring disease activity in multiple sclerosis using serum neurofilament light protein. Neurology 89(22):2230–2237 4. Khalil M, Teunissen CE, Otto M et al (2018) Neurofilaments as biomarkers in neurological disorders. Nat Rev Neurol 14(10):577–589 5. Sandelius Å, Zetterberg H, Blennow K et al (2018) Plasma neurofilament light chain concentration in the inherited peripheral neuropathies. Neurology 90(6):e518–e524 6. Kapoor M, Foiani M, Heslegrave A et al (2019) Plasma neurofilament light chain concentration is increased and correlates with the severity of neuropathy in hereditary transthyretin amyloidosis. J Peripher Nerv Syst 24(4):314–319 7. Mariotto S, Farinazzo A, Magliozzi R et al (2018) Serum and cerebrospinal neurofilament light chain levels in patients with acquired peripheral neuropathies. J Peripher Nerv Syst 23(3):174–177 8. Bischof A, Manigold T, Barro C et al (2018) Serum neurofilament light chain: a biomarker of neuronal injury in vasculitic neuropathy. Ann Rheum Dis 77(7):1093–1094 9. Park SB, Goldstein D, Krishnan AV et al (2013) Chemotherapy-induced peripheral neurotoxicity: a critical analysis. CA Cancer J Clin 63 (6):419–437 10. Wolf S, Barton D, Kottschade L et al (2008) Chemotherapy-induced peripheral neuropathy: prevention and treatment strategies. Eur J Cancer 44(11):1507–1515 11. Carozzi VA, Canta A, Chiorazzi A (2015) Chemotherapy-induced peripheral

neuropathy: what do we know about mechanisms? Neurosci Lett 596:90–107 12. Cavaletti G, Marmiroli P (2015) Chemotherapy-induced peripheral neurotoxicity. Curr Opin Neurol 28(5):500–507 13. Cavaletti G, Frigeni B, Lanzani F et al (2010) Chemotherapy-induced peripheral neurotoxicity assessment: a critical revision of the currently available tools. Eur J Cancer 46 (3):479–494 14. Cornblath DR, Chaudhry V, Carter K et al (1999) Total neuropathy score: validation and reliability study. Neurology 53(8):1660–1664 15. Themistocleous AC, Ramirez JD, Serra J et al (2014) The clinical approach to small fibre neuropathy and painful channelopathy. Pract Neurol 14:368–379 16. Svilpauskaite J, Truffert A, Vaiciene N et al (2006) Electrophysiology of small peripheral nerve fibers in man. A study using the cutaneous silent period. Medicina (Kaunas) 42 (4):300–313 17. Meregalli C, Fumagalli G, Alberti P et al (2018) Neurofilament light chain as disease biomarker in a rodent model of chemotherapy induced peripheral neuropathy. Exp Neurol 307:129–132 18. Meregalli C, Fumagalli G, Alberti P et al (2020) Neurofilament light chain: a specific serum biomarker of axonal damage severity in rat models of chemotherapy-induced peripheral neurotoxicity. Arch Toxicol. https://doi. org/10.1007/s00204-020-02755-w 19. Petzold A (2005) Neurofilament phosphoforms: surrogate markers for axonal injury, degeneration and loss. J Neurol Sci 233 (1–2):183–198 20. Norgren N, Rosengren L, Stigbrand T (2003) Elevated neurofilament levels in neurological diseases. Brain Res 987(1):25–31 21. Kuhle J, Nourbakhsh B, Grant D et al (2017) Serum neurofilament is associated with

NfL as a Biomarker for Chemotherapy-Induced Peripheral Neurotoxicity (CIPN) progression of brain atrophy and disability in early MS. Neurology 88(9):826–831 22. Rissin DM, Kan CW, Campbell TG et al (2010) Single-molecule enzyme-linked immunosorbent assay detects serum proteins at subfemtomolar concentrations. Nat Biotechnol 28 (6):595–599 23. Gissle´n M, Price RW, Andreasson U et al (2015) Plasma concentration of the neurofilament light protein (NFL) is a biomarker of CNS injury in HIV infection: a cross-sectional study. EBioMedicine 3:135–140

275

24. Rohrer JD, Woollacott IO, Dick KM et al (2016) Serum neurofilament light chain protein is a measure of disease intensity in frontotemporal dementia. Neurology 87 (13):1329–1336 25. Kuhle J, Barro C, Andreasson U et al (2016) Comparison of three analytical platforms for quantification of the neurofilament light chain in blood samples: ELISA, electrochemiluminescence immunoassay and Simoa. Clin Chem Lab Med 54(10):1655–1661

Chapter 13 Assessing Neurotoxicant-Induced Inflammation in the Central Nervous System: Cytokine mRNA with Immunostaining of Microglia Morphology Christopher A. McPherson and G. Jean Harry Abstract Inflammation occurs as a normal response of the organism to harmful stimuli such as microbial pathogens, irritants, or toxic cellular components that result from injury and trauma. It serves as a balanced process of pro- and anti-inflammatory responses to maintain normal tissue. The increased role of inflammation in the manifestation of neurotoxicity, whether directly induced or the result of pathological changes, has led to assessments of inflammatory factors within models of environmental exposures. Within the brain, an inflammatory response can be elicited from the resident central nervous system (CNS) glia (microglia and astrocytes) but can also be influenced by endothelial cells and peripherally derived immune cells depending on the nature of the insult, chemical-induced insults. There is a complex and dynamic response in the brain to regulate the inflammatory process. This chapter outlines methods to assess occurrence of a neuroinflammatory response with examination of mRNA levels for pro-inflammatory cytokines and receptors by qRT-PCR, combined with immunocytochemical staining for resident microglia immune cells and their morphological assessment. Analysis of the endpoints described in this chapter provides a framework to assess chemical-induced inflammation in the CNS. Key words Microglia, Pro-inflammatory cytokines, TNF, IL-6, Neuroinflammation

1

Introduction Inflammation occurs as a normal response of the organism to harmful stimuli such as microbial pathogens, irritants, or toxic cellular components that result from injury and trauma. Within the brain, an inflammatory response can be elicited from the resident CNS glia (microglia and astrocytes), endothelial cells, and peripherally derived immune cells. These cells mediate pro- and anti-inflammatory responses and are major sources of soluble molecules, cytokines, chemokines, hormones, and neuropeptides. These small molecules and membrane receptors provide cells with the tools to sense, process, and relay physiological signals beyond their canonical roles. A complex series of immune-like reactions

Jordi Llorens and Marta Barenys (eds.), Experimental Neurotoxicology Methods, Neuromethods, vol. 172, https://doi.org/10.1007/978-1-0716-1637-6_13, © Springer Science+Business Media, LLC, part of Springer Nature 2021

277

278

Christopher A. McPherson and G. Jean Harry

are initiated to neutralize invading pathogens, repair injured tissues, and promote wound healing. In all cases, inflammation represents a coordinated process to serve a specific goal to restore tissue homeostasis. Such reactions include the clearance of cellular debris, secretion of neurotrophic factors, secretion of cytokines, and activation of proteases for matrix remodeling. In this framework, inflammation can be viewed as a complicated series of local immune responses to deal with a threat to the microenvironment. The appropriate regulation of the initial cellular response to tissue damage facilitates recovery while uncontrolled neuroinflammation can induce secondary injury. Within the area of neurotoxicity assessment, neuroinflammation can represent an underlying stress placed upon the system, a cellular process associated with cell injury and repair, as well as a possible mode of action to initiate cell death or exacerbate/prolong injury. This chapter is not intended to serve as a review of the components of neuroinflammation for which there is an extensive literature rather, it will present points of consideration when designing experiments to evaluate exposure-related neuroinflammation and provide methods for examining in vivo morphological responses of microglia and quantification of mRNA expression of inflammatory markers. Many of the general concepts and approaches can be applied to cells in culture. 1.1

Cytokines

Neuroinflammation is closely linked with the induction and presence of inflammatory factors including pro-inflammatory cytokines such as tumor necrosis factor (TNF), interleukin 1 (IL-1), and IL-6 [1]. Elevations in such factors can represent a response to adjacent cellular damage, direct action on immune cells, cell interactions, and infiltration of peripheral immune cells [2]. Upon activation of transcription factors or inhibition of negative transcription regulators, mRNA synthesis is initiated for the eventual production and secretion of pro-inflammatory cytokines. Increased expression of cytokines occurs transcriptionally, post-transcriptionally, translationally, and through the conversion of latent precursors to biologically active protein [3–9]. Translational control offers a strategic advantage to these cells, allowing the use of pre-existing mRNAs to bypass the lengthy nuclear control mechanisms (e.g., transcription, splicing, and transport); additionally, it provides for reversibility through modifications of the regulatory intermediates. Combined, these features allow rapid activation or termination of synthesis of a specific protein or group of proteins required for the inflammatory process. Cytokines can be secreted upon complete intracellular processing while others are stored intracellularly and require an additional stimulus to trigger protein secretion. There is additional evidence demonstrating that, while most cytokines are secreted from the cell in a biologically active form for some, the latent form can be secreted, requiring additional extracellular processing for biological

Assessing Neurotoxicant-Induced Inflammation in the Central Nervous System. . .

279

activity. A biologically active cytokine can bind to a specific membrane receptor, a carrier protein, or a soluble receptor. This allows for induction and regulation of the inflammatory response often in an autocrine/paracrine fashion. For the pro-inflammatory cytokine, tumor necrosis factor (TNF) signaling, upon release of mature biologically active protein, TNF-α interacts with two different receptors, designated TNFR1 and TNFR2, which are differentially expressed on cells and tissues and initiate both independent and overlapping signal transduction pathways, leading to multiple cellular responses [10]. In addition to membrane-associated receptors, soluble forms of some cytokine receptors can be generated in response to cell activation such as soluble TNF receptors that bind released TNF and block its biological activity. The interleukin-1 (IL-1) family of cytokines and receptors broadly affects a broad range of inflammatory responses [11]. IL-1α and IL-1β are of primary interest in the nervous system. The family of IL-1 receptors contains pro- and anti-inflammatory receptors including IL-1R1 that binds IL-1α, IL-1β, and IL-1 receptor antagonist (IL-1Ra) [12]. The association between cytokine production and the concurrent upregulation of receptors allows for the opportunity to use mRNA levels to examine biological activity of a specific cytokine by measuring both the cytokine and the cytokine receptor. This would then allow for examination of downstream signaling events as a result of changes observed in cytokine mRNA levels. 1.2 Consideration of the Source of Inflammatory Factors

Under healthy conditions, microglia are the only immune cell type present in the CNS parenchyma yet there are various immune regulatory cells such as macrophages and dendritic cells present in the CNS-adjoining tissues [13, 14]. The CNS is mechanically separated from the circulation by the blood–brain barrier (BBB). This separation influences immune responses by excluding many peripherally derived innate and adaptive immune cells and inflammatory molecules. However, infiltrating cells significantly contribute to any neuroinflammatory response following disruption of the BBB, as can occur with physical injury or high levels of inflammation. While a predominant interest is in the response of resident nervous system immune cells, microglia, the possible contribution from infiltrating blood-borne cells or circulating factors requires consideration. Induction of a peripheral inflammatory response and elevations of inflammatory factors in the blood can influence either a localized response along the vascular wall or actually infiltrate into the brain parenchyma. Thus, the contribution from peripheral immune cells must be considered with any experimental design and interpretation of the resulting data. The responses of microglia versus peripheral macrophages display distinct properties under polarization conditions and may play different roles in the inflamed CNS [15, 16]. Figure 1 shows an example of how the blood compartment can significantly influence the data outcome.

280

Christopher A. McPherson and G. Jean Harry

Saline Perfused

20 Fold Change mRNA

Non-Perfused 15

10

5

0 Tnfa

Il6

Fig. 1 Adult male mice were exposed to Chemical X for 3 months. At 3 months following exposure, Tnfa and Il6 mRNA expression was analyzed from the hippocampus of saline perfused vs non-perfused animals. mRNA expression for Tnfa and Il6 was significantly higher in the non-perfused animals compared to saline perfused animals (*p < 0.05). These data demonstrate the importance of considering the contribution of peripheral blood compartment 1.2.1 Peripheral

The BBB mechanically separates the CNS from the circulation by the presence of specialized endothelial cells tightly attached to each other via tight junctions and adherens junctions. As such, it excludes plasma proteins and many of the peripherally derived innate and adaptive immune cells and associated inflammatory molecules [17, 18]. Thus, it can be considered that the BBB actively contributes to the immune response of the CNS not only by regulating entry of peripheral factors but also with signaling upon stimulation of inflammatory cells along the vascular wall and perivascular microglia. With physical injury to the BBB or highly inflamed CNS parenchyma, blood-borne monocytes can enter the brain parenchyma and contribute to the inflammatory response and pathogenic process [19, 20].

1.2.2 Central Nervous System

In the CNS, microglia serve as the primary resident innate immune cell population [21]. Depending on the anatomical region, microglia account for between 0.5 and 16% of the total cell population in the human brain [22] and 5–12% in the mouse brain [23]. Microglia are unique from peripheral monocytes in that they develop from myeloid progenitors from the embryonic yolk sac [24, 25]. Microglia migrate into the embryonic CNS where they proliferate to populate the brain without further contribution of peripheral progenitors to the resident microglia pool [26, 27]. This developmental process establishes a life-long population of resident microglial cells which is maintained by a limited self-renewal [28– 30]. This also makes the early establishment of the microglia

Assessing Neurotoxicant-Induced Inflammation in the Central Nervous System. . .

281

population a critical event in development that can have long-term consequences. Thus, one can consider that this low turnover rate renders microglia susceptible to the effects of environmental exposure, age, injury, stress, or any combination. Microglia display morphological heterogeneity and differential density across brain regions [23, 31]. They dynamically and actively sense their microenvironment and perform crucial physiological functions for tissue development, architecture refinement, and tissue remodeling/ repair [32, 33]. In their quiescent surveillance phenotype, microglia display fine mobile processes extending into the surrounding microenvironment [32]. The initial concept of a stereotypic response of microglia with limited response variability has been replaced by data showing that microglia respond with a variety of different morphological changes as a result of integrating multifarious inputs and represent a diverse spectrum of responses that do not fall within a dichotomy of surveying or activated [33– 35]. When microglia sense pathogens, dying cells, debris, or aberrant proteins, they adjust to their environment which often manifests as a phenotypic change including morphological changes and increase in phagocytic capability. As damage is resolved, the correlated pro-inflammatory response diminishes. 1.2.3 Distinguishing Cell Origin

While it is enticing to consider any specific marker as unique for either resident microglia or peripheral monocytes/macrophages, such confirmation of the cellular source requires additional approaches. One approach relies on flow cytometry to discriminate between microglia [CD11b+ and CD45 low] and peripheral macrophages [CD11b+ and CD45 high] [36]. A more confirmative method used to identify infiltrating blood-borne monocytes and discriminate from resident microglia relies on bone marrow chimera animals to track fluorescently tagged bone marrow-derived cells [37]. While microglial gene expression shares homology with tissue-resident macrophages, it maintains a unique gene expression signature [38, 39].

1.3 Microglia Receptors

The distinct features associated with a neuroinflammatory response will impact the design and interpretation of experimental models to assess effects within a neurotoxicology framework. In a sterile inflammatory response, immune cells are activated in the absence of microbial compounds by endogenous molecules called dangerassociated molecular patterns (DAMPs) [40]. These damage signals can activate immune cells through pattern recognition receptors (PPR) including: Toll-like receptors (TLRs), nucleotidebinding domain leucine-rich repeat containing proteins (also known as NOD-like receptors, NLRs), and Rig1-like receptors (RLRs). The engagement of these receptors results in the induction of specific pathways and release of cytokines that contribute to resolving injury. Molecules that function as DAMPs include nucleic

282

Christopher A. McPherson and G. Jean Harry

acids, lipids, proteins that immune cells do not normally see until they are released or unmasked with cells death occurring due to tissue injury. It is thought that receptors on microglia can act as molecular switches to control microglial responses [33, 41]. Microglia express age and sex-dependent P1 receptors for nucleosides and monophosphate containing nucleotides (adenosine and AMP) and most of the receptors of the P2 family for di- and tri-phosphate containing nucleotides (e.g., ADP, ATP) that can regulate both pro-inflammatory and anti-inflammatory functions [42]. 1.4

Cell Polarization

In general, the macrophage inflammatory response is characterized based on the nature of the activating stimulus and the resulting production of factors which can span a vast spectrum [34, 43– 45]. A diverse array of inflammatory stimuli has been used alone or in combinations to generate polarized macrophage populations in vitro and in vivo. While a classification of M1 and M2 for a dichotomy of responses has taken hold in the neuroinflammation literature, this is considered an inappropriate and simplified terminology. Rather, it is recommended that the stimulus inducing the response should be identified [45]. Examples include M[LPS], m [IFNγ], M[LPS + IFNγ] to elicit release of pro-inflammatory factors and m[IL-4], M[Il-10], and M[IL-13] to elicit the release of anti-inflammatory factors. Stimulation of these pathways lead to production of pro-inflammatory cytokines (e.g., interferon gamma (IFNγ), IL-12, TNF-α, IL-6, and IL-1β), chemokines (CCL2, CXCL10, and CXCL11), and antigen presentation molecules, such as major histocompatibility complex (MHC) or antiinflammatory cytokines (e.g., IL-4, IL-10, IL-13, and transforming growth factor beta, (TGF-β)) as well as arginase-1 (Arg-1), CD206, and Chitinase-3-like-3 (Ym-1 in rodents). As research on polarization advances, it is becoming clear that reliance on a limited number of factors for either polarization state is not recommended or reflective of the biology.

1.5

Age

Microglia originate from a primitive monocyte population derived from the yolk sac during a defined window of time before vascularization or definitive hematopoiesis and in rodents, they are present as early as embryonic day 8 [28, 29]. The cells demonstrate a structural maturation that starts from a more rounded morphology in the early stages of development followed by a progressive complexity of cell processes [46]. Many of the pro-inflammatory and anti-inflammatory factors also show a pattern over the course of development followed by a shift with aging [47]. Thus, when designing a study or interpreting data from a developmental or aging study, reliance upon literature derived from the adult animal is not recommended given the multiple and shifting roles for cytokine signaling. Also, with the dynamic nature of brain development, including the high level of cell death and remodeling, any

Assessing Neurotoxicant-Induced Inflammation in the Central Nervous System. . .

283

elevation in phagocytic action of microglia may reflect the normal high demands placed on the cells for clearance of cellular material. With aging, a decreased phagocytic capability would be considered detrimental. Over the life span, microglia continue to change. Morphologically, microglia in the aged brain display cytoplasmic structures showing excessive beading and spheroid swellings that may be reflective of dystrophy and senescence [48]. The observed decrease in process motility may compromise surveillance features and responsivity [49, 50]. In addition, a slight elevation in the basal level of pro-inflammatory cytokines and a diminished ability to recruit an anti-inflammatory response is observed. Thus, as these limited examples demonstrate, it is critical that investigators ensure an understanding of the related biological processes as they apply to the age under study. 1.6 Time Interval for Assessment

An acute or short-term exposure regimen requires consideration of when to assess for an associated neuroinflammatory response. The sequence of biological events required to initiate a neuroinflammatory or microglia response would require time for the effect to occur and reach an adequate level of detection. For example, an acute response might rely on chemoattractant signaling molecules such as the chemokines, transcription factor modifications allowing for cytokine production, and receptor activation for a downstream biological effect. Many pro-inflammatory cytokines display an autocrine and paracrine signaling and thus, self-regulate expression. In addition, many cytokines share an ATTTA sequence in their 30 end that allows for rapid mRNA degradation [51]. Thus, depending on the time of assessment, relative to the pro-inflammatory response, different stages of the response may be required.

1.7 Quantitative Real-Time PCR for mRNA Levels of Inflammatory Markers In Vivo

Quantitative reverse transcription-polymerase chain reaction, qRT-PCR, uses reporter dyes to detect PCR products in real time. The assay relies on measuring the increase in fluorescent signal, which is proportional to the amount of DNA produced during each PCR cycle. Individual reactions are characterized by the PCR cycle at which the fluorescence first rises above a defined background level, a parameter known as the threshold cycle (Ct). The more gene target there is in the starting material the lower the Ct. This correlation between fluorescence and the amount of amplified product allows for a quantitated assessment of target mRNA expression. Detection of qRT-PCR products can be accomplished with either SYBR™ Green dye or TaqMan® probes. SYBR™ Green dye fluoresces when bound to the double-stranded PCR product amplicons from investigator-designed forward and reverse PCR primers. The primers attach to the anti-sense and sense strand of the target DNA of interest. As the PCR progresses, more amplicons of the target sequence are created which in turn bind more SYBR™ Green, resulting in increased fluorescent intensity that is directly

284

Christopher A. McPherson and G. Jean Harry

Table 1 Advantages and limitations of SYBR™Green and TaqMan® probes Detection method ™

SYBR Green

Advantages l

l

Primers can be designed for any double-stranded DNA sequence Does not require probes, lower assay costs

Limitations l

l

l

l

TaqMan®

l

l l l

Specific hybridization between target and probe is required to generate fluorescent signal High sensitivity for low copy targets Dissociation curves are not necessary Predesigned and validated probes are commercially available

l

l

SYBR Green binds to all double-stranded DNA generated in the reaction and may generate false positives Sensitivity varies depending on primer design Time-consuming primer optimization is required for assays A dissociation curve analysis PCR product is required at the end of each assay to ensure specificity of primer binding, lengthening the assay time Probes cost more than SYBR Green Primers A different probe needs to be generated for each target sequence of interest

Adapted from: http://tools.thermofisher.com/content/sfs/manuals/cms_083618.pdf

proportional to the amount of target gene in the sample. Important optimization steps have been described [52, 53] and must be undertaken when SYBR™ Green probes are used. All of these optimization steps are done to ensure sensitivity, specificity, reproducibility, and a large dynamic range of the PCR reaction. They are briefly outlined in Subheadings 3.5.4.1 and 3.5.4.2. TaqMan® probes contain a reporter dye (FAM or TAMRA) linked to the 50 end of the probe, and a non-fluorescent quencher (NFQ) at the 30 end of the probe. While intact the TaqMan® fluorescent reporter on the 50 end is quenched by the NFQ on the 30 end. When the probe binds to the target sequence, it is cleaved by Taq DNA polymerase during primer extension. This cleavage separates the quencher resulting in an increase in the reporter dye. With each PCR cycle probes bind and are cleaved, increasing the fluorescent signal intensity. The resulting increased fluorescent intensity is directly proportional to the amount of target sequence in the sample. The advantages and disadvantages of each detection chemistry (Table 1) should be considered when designing experiments. 1.7.1 Target Gene Primers

The assay should be specifically designed for each target gene to identify all transcripts of interest including splice variants while discriminating between closely related family members. This

Assessing Neurotoxicant-Induced Inflammation in the Central Nervous System. . .

285

protocol outlines the use of “off the shelf” TaqMan® assays that are confirmed to be specific to the genes of interest allowing for high specificity detection of low copy number inflammatory molecules in the CNS. It is also important to ensure that the assay has an amplification efficiency close to 100% in the tested system. A less efficient assay can result in reduced sensitivity and reduced dynamic range, limiting the ability to detect low abundance cytokine transcripts. The quantitation method should be determined. The procedures outlined in this chapter have been validated for use of either the comparative Ct (2ΔΔCt) or the relative standard curve method. The 2ΔΔCt method is a relative quantitation method used to analyze fold changes of gene expression in a given sample relative to a reference sample (i.e., untreated control), using the mathematical model first described by Livak and Schmittgen [52]. The 2ΔΔCt method is useful to compare levels in treated vs untreated samples within a study. The relative standard curve method offers more direct analysis options when a more complex experimental design is needed such as measurement expression levels in samples treated with a compound under different experimental conditions. This is especially valuable under conditions where any prior manipulation or exposure alters the baseline expression level of a gene prior to any secondary manipulation. Using the standard curve method allows for a more refined experimental design and statistical analysis. The relative standard curve method uses a dilution series created from a positive control sample run with both the target and endogenous control gene. The standard curve can be generated from any known biological source of the gene but needs to generate a curve for which the experimental samples fall within the linear portion. For all experimental samples, a relative quantity is determined from this dilution series [53]. 1.7.2 Selection of Housekeeping Gene

Regardless of the quantitation method used, selection of a valid endogenous control to normalize for RNA sampling is critical to avoid misinterpretation of results. Suitable endogenous controls include housekeeping genes such as GAPDH, β-actin, or 18S ribosomal RNA (18S rRNA). It is imperative that expression of the endogenous control gene used in the qRT-PCR reaction does not change with the experimental manipulation. Subtle changes in the level of a high-copy housekeeping gene that may not appear significant can significantly influence interrogating low copy number cytokine mRNA transcripts. The fold change of endogenous housekeeping genes following treatment can be calculated using the 2Ct method (Table 2) [53]. It is suggested that, for any new experimental manipulation, selection of the housekeeping gene includes comparing of more than one gene as well as the impact of any subtle changes on target gene expression level is evaluated.

2(17.5) ¼ 5.36E06

2(17.4) ¼ 5.64E06

2(17.6) ¼ 4.92E06

2(17.9) ¼ 4.18E06

2(16.9) ¼ 8.01E06

Mean ¼ 5.63E06

2(16.6) ¼ 8.01E06

2(16.6) ¼ 9.78E06

2(17.8) ¼ 4.25E06

2(18.0) ¼ 3.95E06

2(17.7) ¼ 4.59E06

Mean ¼ 5.23E06 -Ct

Mean ¼ 8.06E07

2(20.5) ¼ 6.63E06

2(20.7) ¼ 5.97E06

2(20.6) ¼ 6.12E06

2(20.6) ¼ 6.12E06

2(19.3) ¼ 1.54E06

β-actin Chemical X

-Ct

Mean ¼ 1.06E06

2(18.9) ¼ 2.09E06

2(19.7) ¼ 1.21E06

2(19.3) ¼ 1.51E06

2(19.3) ¼ 1.51E06

2(19.2) ¼ 1.76E06

Vehicle

Mean ¼ 5.23E06

2(20.8) ¼ 5.64E06

2(21.1) ¼ 4.39E06

2(20.8) ¼ 5.36E06

2(20.9) ¼ 5.28E06

2(19.6) ¼ 1.22E06

18S rRNA Chemical X

Gene expression of endogenous controls can be determined using the 2 method. First calculate the mean 2 for each housekeeping gene Next calculate the fold change of each housekeeping gene GAPDH: 5.23E06/5.63E06 ¼ 9.29E01; 1/9.29E01 ¼ 1.08 fold change β-actin: 8.06E07/1.06E06 ¼ 5.03E01; 1/5.03E01 ¼ 2.0 fold change 18s rRNA: 5.23E06/5.63E06 ¼ 1.27 fold change It is recommended not to use a housekeeping gene that changes over 1.4-fold under experimental conditions [43, 44]

Vehicle

GAPDH Chemical X

Table 2 Validation of endogenous housekeeping gene

Mean ¼ 5.63E06

2(20.5) ¼ 6.67E06

2(21.4) ¼ 3.71E06

2(20.9) ¼ 5.25E06

2(21.1) ¼ 4.42E06

2(20.7) ¼ 5.68E06

Vehicle

286 Christopher A. McPherson and G. Jean Harry

Assessing Neurotoxicant-Induced Inflammation in the Central Nervous System. . .

X al ic m he C

C

he

m

ic

ic

al

le

X

0.0

Ve h

ic m he C

Ve h

ic

al

le

X

0.0

0.5

ic le

0.0

0.5

1.0

Ve h

0.5

1.0

18S rRNA

1.5

Fold Change mRNA

1.0

-actin

1.5

Gapdh Fold Change mRNA

Fold Change mRNA

1.5

287

Fig. 2 Three housekeeping genes (Gapdh, b-actin, and 18 s rrna) were examined for fold change following exposure to Chemical X, using the 2-Ct method (Table 2). The fold change for Gapdh, 18 s rrna, and b-actin were 1.0, 1.2, and 2.0, respectively. When Gapdh and 18 s rrna were each used as the endogenous controls in the 2ΔΔCt calculation to compare differences with vehicle controls, a significant decrease in Tnfa mRNA level was observed following exposure to Chemical X (*p < 0.05). In contrast, when b-actin was used, no changes in Tnfα mRNA levels were observed. Chemical X reduced b-actin mRNA levels by 2.0-fold and was not considered a suitable internal control gene

An example of how changing housekeeping gene expression may affect interpretation of the pro-inflammatory cytokine TNF-α mRNA levels as determined by the 2ΔΔCt method is provided (Fig. 2). 1.8 Considerations for an Experimental Approach

It is highly recommended that a spectrum of markers is employed to determine neuroinflammation. Of additional interest are to determine if any change is a result of chemical exposure, if it occur in the absence of neuropathology, and if it is sufficient to initiate a biologically active inflammatory response. These markers should reflect not only the initial stage but also confirm a biological downstream activation. This could be a combination of mRNA levels for pro-inflammatory cytokines, associated receptors, receptor antagonists, and anti-inflammatory cytokines. The selection of genes may reflect whether or not the system is capable of responding to the inflammatory process for down-regulation. Depending on the model system, one may want to consider additional markers that reflect resolution and repair. Combining neuroanatomical assessments with molecular or biochemical assessments lends support for any interpretation. A change in microglia morphology as a result of damage in the surrounding environment is characterized by increased soma size and the retraction of elongated fine processes to shorter, coarser cytoplasmic processes displaying a bushy appearance. This

288

Christopher A. McPherson and G. Jean Harry

Fig. 3 Representative images of microglial response to chemical-induced neuronal cell death in the hippocampus of adult mice. Adult male CD-1 mice received a single intraperitoneal injection of trimethyltin hydroxide (TMT; 2.3 mg/kg), (a) Iba-1+ microglia cells (green), and nuclear DAPI (blue) in controls and with neuronal death at 72 h post-injection. In the control hippocampus, Iba-1+ cells displayed thin elongated processes with distal ramifications. With neuronal death, Iba-1+ cells differentiated into a rounded amoeboid morphology suggesting a high level of phagocytic activity to remove dead neurons in the dentate granule region and in the CA pyramidal region cells with enlarged bushy cells with retracted processes were observed in areas not necessarily associated with neuronal death. Cells are also present that do not appear to shift significantly from controls. (b) Representative rating scale that can be used for discriminating microglia morphology based on cell soma shape, and orientation, density, and complexity of processes

morphology can further shift to fully ameboid morphology depending on the nature and severity of the insult (Fig. 3). The response in a developing animal requires additional considerations given the maturation of resident microglia with regard not only to morphology but also to cytokine production and the functions of such cytokines. It is recommended that the morphology of the cell be evaluated rather than to either count number of cells or simply assign a classification to the cell response A number of quantitative and semi-quantitative approaches have been developed for this purpose [54, 55].

2

Materials The protocol is focused on the use of rodents (mice, rats). For any use of animal, procedures should be conducted in compliance with all requirements of institutional committees regarding use of animals in research.

Assessing Neurotoxicant-Induced Inflammation in the Central Nervous System. . .

2.1 Whole-Body Saline Perfusion Equipment and Materials

289

1. Balance to weigh animal. 2. Perfusion pump or 50 mL syringe. 3. Restraint board to restrain body and limbs. 4. Small dissecting scissors. 5. Hemostat to secure perfusion needle. 6. Fatal-Plus® Solution (Vortech Pharmaceuticals, Ltd., 9373). 7. 25 g x ¾00 Winged perfusion needle with 1200 tubing. 8. Sterile 0.9% Saline solution.

2.2

Excision of Brain

1. Scissors for decapitation. 2. Small surgical scissors for cutting scalp. 3. Small animal rongeurs or sturdy forceps to remove skull. 4. Flat edge weighing spatula to remove brain from skull. 5. Single edge razor blade to block brain.

2.3

Tissue Collection

2.3.1 Frozen Tissue

1. Flat surface on dry ice (see Note 1). 2. Labeled RNAse-free microcentrifuge tubes (1.5–1.8 mL) for frozen tissue storage. 3. RNaseZap® RNase decontamination solution (Thermo Fisher, AM9780).

2.3.2 Fixed Tissue

1. Labeled flat bottom containers. 2. 0.1 M phosphate buffer (PB) pH 7.4 (a) 77.4 mL 1 M Na2HPO4 (Sigma, S5136). (b) 22.6 mL 1 M NaH2PO4 (Sigma, S5011). (c) 900 mL dH2O. 3. 4% weight/volume (w/v) paraformaldehyde solution (see Note 2) (a) 0.4 g Paraformaldehyde, EM grade (Ted Pella, 18501). (b) 0.1 M phosphate buffer (bring to 10 mL using a volumetric flask). 4. 15% (w/v) sucrose solution (a) 150 g sucrose. (b) 0.1 M PB (bring to 1 L using a volumetric flask). 5. 30% (w/v) sucrose solution (a) 300 g sucrose. (b) 0.1 M PB (bring to 1 L using a volumetric flask). 6. Tissue-Tek® O.C.T. Compound (Sakura, 4583). 7. Tissue-Tek® Cryomold (Andwin Scientific, 4557).

290

2.4

Christopher A. McPherson and G. Jean Harry

Cryosectioning

1. Cryostat (see Note 3). 2. Disposable cyrostat blades (Thermo Fisher, 3051835). 3. Fine artist paintbrush (#3) for handling sections to minimize damage to sections. 4. 2.0 mL cryovials (Nunc, 368632) 5. 6-well tissue culture plate or equivalent 6. 0.1 M PB 7. FD Section Storage Solution® (FD NeuroTechnologies, Inc., PC101) (see Note 4).

2.5 Immunofluorescent Staining

1. 24-well tissue culture plate or equivalent. 2. Fisherbrand™ Superfrost™ Plus Microscope Slides (Fisher, 22-037-246). 3. Coverslips. 4. Fine paint brush. 5. Clear nail polish. 6. Fluorescent microscope. 7. Blocking Solution (a) 0.1 M PB. (b) 10% Normal Goat Serum (Gemini Bio, 100-109). (c) 0.3% Triton X-100. 8. Anti Iba1, Rabbit polyclonal (Wako 019-19741) (see Note 5). 9. Goat anti-rabbit IgG Alexa Fluor488 conjugate (Invitrogen, A-11008). 10. 40 ,6-diamidino-2-phenylindole, 5 mg/mL stock solution (DAPI; Invitrogen D1306) (see Note 6) 11. Dissolve the contents of one vial of DAPI (10 mg) in 2 mL of dH2O. 12. ProlongGold™ Antifade Mountant (Invitrogen, P36930).

2.6

qRT-PCR

1. Latex gloves. 2. Hand-held tissue homogenizer with polytron probe (see Note 7). 3. Round bottom 2 mL microcentrifuge tubes (see Note 8). 4. RNase-free Microcentrifuge tubes (1.5 mL). 5. Vortex. 6. PCR machine. 7. Spectrophotometer. 8. Pipette and tips.

Assessing Neurotoxicant-Induced Inflammation in the Central Nervous System. . .

291

9. Optical 96-well or 384-well PCR plate (Applied Biosystems, 4326813 or 4309849). 10. Optical adhesive film (Applied Biosystems, 4360954). 11. Multichannel pipet 0.5–10 μL and 5–100 μL. 12. TRIzol™ (Invitrogen, 15596026) (see Note 9). 13. 70% ethanol 14. 100% Isopropanol 15. Fresh chloroform (see Note 10). 16. RNAse-Free H2O. 17. Superscript II Reverse Transcriptase (Invitrogen, 18064014). 18. Oligo dT primer (Applied Biosystems, N8080128). 19. Oligonucleotides, 18420788).

10

mM

DNTP

mix

(Invitrogen,

20. TaqMan® Universal PCR master mix (Applied Biosytstems, 4305719). 21. TaqMan® Gene Expression Assays FAM (Applied Biosystems, 4331182). 22. Power SYBR™ Green master mix (Applied Biosystems, 4368706).

3

Methods

3.1 Whole-Body Saline Perfusion

1. Weigh animal. 2. Deeply anesthetize animal with a single intraperitoneal injection 100 mg/kg body weight Fatal-Plus® Solution. 3. Once anesthesia has taken effect secure the animal on a dissection stage with each limb immobilized. 4. Using a scalpel make an incision along the underside of the animal to expose the chest cavity. 5. Using scalpel or small surgical scissors open the chest cavity without injuring the lungs or heart. 6. Insert the perfusion needle (winged or blunt edge needle shaft) into the left ventricle of the heart and stabilize then perform a small incision on the right atrium. 7. Perfuse the mouse with a minimum of 10 mL (mouse) or 30 mL (rat) of cold 0.9% saline using a syringe or diastolic perfusion pump attached to the needle inserted into the left ventricle of the heart. Saline should be perfused at a rate of 1 mL per minute. 8. Following perfusion, decapitate the head and carefully extract the brain from the skull.

292

3.2

Christopher A. McPherson and G. Jean Harry

Tissue Collection

3.2.1 Histology and mRNA from Each Brain to Allow for Assessments in Each Animal (See Note 11)

1. Place the brain on a flat surface with cortical side up and cut with a single or double edge razor blade along the midline of the brain hemispheres to bisect the brain into two hemispheres. Prefer a straight edge of a razor blade to a scalpel blade to facilitate straight even cut. 2. To fix tissue, place one hemisphere in flat bottom container (e.g., specimen jar) containing fresh 4% paraformaldehyde/ PBS (approximately 50 mL/g tissue weight). 3. Ensure that the cut face of the tissue lies on the flat surface to maintain accurate cytoarchitecture. 4. Allow brains to fix in 4% paraformaldehyde/PBS for 18 h at 4  C. 5. If additional blocking of brain is required for specific regional sectioning, this is conducted only after tissue fixation. 6. Proceed to Cryopreservation Subheading 3.3.1. 7. From the non-fixed contralateral extracted hemisphere, subdissect any region of interest (e.g., freehand, micro-punch), 8. Immediately place tissue on a pre-chilled flat metal surface on a bed of dry ice and allow tissues to freeze. 9. Store froze tissue in microcentrifuge tubes at 80  C. Frozen tissue will remain loose in the tube during storage and if adhered to the tube wall that will indicate a thaw/freeze during storage and degradation.

3.3

Cryosectioning

3.3.1 Cryopreservation

1. Transfer brain to a flat bottom specimen jar containing 15% w/v sucrose solution in 0.1 M PB (approx. 30 mL/g tissue weight), incubate at 4  C until the brain sinks, approximately 3 days. 2. Transfer brain to specimen jar containing equivalent volume of 30% (w/v) sucrose solution in 0.1 M PB, incubate at 4  C until the brain sinks. 3. Remove cryo-preserved brains from the sucrose. 4. Place tissue with the cutting face down into a cryomold, add OCT to each tissue cassette sufficient to completely cover the brain. 5. Immediately freeze brain in cryomold on dry ice, ensuring maintaining correct architecture, store at 20  C.

3.3.2 Cryosectioning

1. Cool cryostat to 20  C before beginning. 2. Remove brain from cryomold and place in cryostat for 15 min to equilibrate to temperature. 3. Mount brain on chuck with OCT compound and allow to freeze for 15 min.

Assessing Neurotoxicant-Induced Inflammation in the Central Nervous System. . .

293

4. Mount chuck into the specimen holder, orient specimen to desired position using adjustable screws on the holder. 5. Insert disposable blade into the knife holder adjust knife angle to 5–10 . 6. Set section thickness to 10 μm. 7. Trim the block to expose the brain region of interest. This is done by carefully taking 10 μm sections through the block. Initial sections will be OCT only and eventually will contain brain samples. If the trimming is not collecting a full uniform section of the sample, adjust the specimen holder until a full section is obtained. 8. Once the block is trimmed set the section thickness to 30 μm. 9. Section and collect tissues using a fine artist paintbrush (#3) into a 6-well plate containing 0.1 M PB (see Note 12). 10. When sectioning is completed, sections can be either processed immediately or gently transferred to cryovial containing section storage solution and stored at 20  C until processing. 3.4 Immunofluorescent Staining 3.4.1 Day 1

1. Add 800 μL 0.1 M PB to the required number of wells of a 24-well tissue culture plate or equivalent. 2. Remove required brain sections from the section storage solution using a fine artist paintbrush and place one section into each well of the 24-well plate containing 0.1 M PB. 3. Allow sections to equilibrate to room temperature (RT) for 15 min. 4. Gently remove fluid and replace with 800 μL 0.1 M PB. 5. Block nonspecific IgG binding in sections by applying 200 μL Blocking Solution with gentle orbital shaking for 2 h at RT. 6. Prepare primary Iba-1 antibody solution by diluting in Blocking Solution. For Wako Chemicals Iba-1 antibody, the recommended dilution is 1:200, this may require optimization dependent on the brain region of interest. Prepare sufficient antibody solution to incubate each section with 200 μL. Store on ice while sections are undergoing blocking (step 5). 7. Aspirate blocking solution using pipet and add 200 μL of primary antibody solution to each section. Allow to incubate at 4  C for 48 h using gentle orbital shaking.

3.4.2 Day 2

1. Gently remove primary antibody solution from the wells. 2. Wash sections 3 with 0.1 M PB, 5 min each. 3. Prepare secondary antibody detection solution by diluting goat anti-rabbit IgG Alexa Fluor488 conjugate 1:5000 in blocking solution.

294

Christopher A. McPherson and G. Jean Harry

4. Add 200 μL secondary antibody solution to each section. Allow to incubate for 2 h at RT with gentle orbital shaking. 5. Gently aspirate secondary antibody solution from wells. 6. Wash sections 3 with 0.1 M PB, 5 min. 7. Prepare DAPI solution by diluting DAPI stock 1:25000 in 0.1 M PB. 8. Add 200 μL DAPI solution to each section allow to incubate at RT for 5 min. 9. Wash sections 3 with 0.1 M PB, 5 min. 10. Using a fine paint brush, gently lift section from well and place on glass microscope slide with adjustments for flat and correct orientation. Remove excess liquid from the slide using a dry laboratory wipe. 11. Apply 1–2 drops of ProLong™ Antifade Mountant to the mounted section. 12. Cover the sections with a coverslip and allow to dry for 24 h at RT in the dark. Drying for 24 h allows for optimal performance of the antifade reagent. 13. Seal stained slides by applying a thin layer of clear nail polish around the outer perimeter of the coverslips; don’t apply sealant over the section. Sealed sections are ready for viewing or storage at 20  C. 3.5

qRT-PCR

3.5.1 RNA Extraction Using TRIzol™ Reagent

1. Maintain your experimental design within the isolation and RNA analysis (see Note 13). 2. Isolate 6 or maximum of 8 samples at a time. 3. Add 1 mL of TRIzol™ Reagent per 50–100 mg to sample in microcentrifuge tube (2.0 mL) (see Note 14). 4. Disrupt subdissected tissue using a polytron tissue homogenizer for 30–40 s. 5. After each tissue, rinse homogenizer probe using 80% ethanol/ RNase-free H2O followed by a rinse with RNase-free H2O. Ensure no residual ethanol on the probe. 6. Centrifuge at 12,000  g at 4  C for 5 min. 7. Transfer the cleared supernatant to a new 1.5 mL microcentrifuge tube. 8. Incubate the homogenized sample for 5 min at room temperature to permit complete dissociation of the nucleoprotein complex. 9. Add 0.2 mL of chloroform per 1 mL of TRIzol™ Reagent used for homogenization. Cap the tube securely. 10. Vortex tube for 15 s.

Assessing Neurotoxicant-Induced Inflammation in the Central Nervous System. . .

295

11. Incubate for 2–3 min at room temperature. 12. Centrifuge the sample at 12,000  g for 15 min at 4  C. Following centrifugation, the mixture separates into a lower red phenol-chloroform phase, an interphase, and a colorless upper aqueous phase. RNA remains in the aqueous phase. 13. Remove the colorless upper aqueous phase of the sample by angling the tube and pipetting the solution out. Avoid drawing any of the interphase or organic layer when removing the aqueous phase and place into a new tube. If material from the interphase or lower phase is drawn up, allow the liquid to partition in the pipette tip and gently push out to discard the unwanted phase and collect the upper phase. 14. Add 0.5 mL of 100% isopropanol to the aqueous phase per 1 mL of TRIzol™. Reagent used for homogenization and incubate at room temperature for 10 min. 15. Place tubes into the centrifuge with a standard orientation to identify where the pellet would form (e.g., plate the hinge of the cap outward and the pellet will form on that side of the tube). 16. Centrifuge the sample at 12,000  g for 10 min at 4  C. Following centrifugation, the RNA should form a gel-like pellet at the bottom of the tube. 17. Gently remove supernatant from the tube, leaving only the RNA pellet. 18. Gently wash the pellet with 1 mL of 75% ethanol. 19. Vortex the samples briefly, then centrifuge the tube at 7500  g for 5 min at 4  C. 20. Air dry the RNA pellet for 5–10 min. It is critical that all ethanol is evaporated from the pellet but if the pellet dries to hard it may be difficult to resuspend and may require a series of freeze/boil to resuspend. 21. With samples of approximately 80–100 mg starting material, resuspend RNA pellet in 30 μL RNAse-free water. For samples derived from starting material Playback control

Tells software to start tracking

>Video window

Shows tracking of larvae

Track smoothing parameters

Removes sources of noise during tracking

ANALYSIS

Reports results of tracking

7

Data profiles

Select trials to report and time Data reported here bins in 2-minute bins

8

Analysis profiles

Endpoints to be calculated

9

Results >Statistics and charts

5

6

>Track visualization

Calculates endpoints Exports data to excel file Tracks (movements) are displayed

DDS (detection determines speed) should be checked Verify that larvae movements are being tracked We use MDM (minimal distance moved) set t o 0.135 cm

Data reported here is distance moved

Larval Zebrafish Locomotor Activity Assay

345

statistically analyze our acclimation period, we still include this as part of our protocol settings and recording. Ethovision allows for the tracking of up to 100 wells (i.e., arenas) with one larva per well. We select all wells (i.e., 96) for tracking and then once tracking is complete, we remove data collected for wells containing dead, unhatched, and abnormal larvae that were identified in Subheading 2.10. Arenas can be subdivided into zones: in zone, distance to zone, and distance to point (as in, distance to the center) can be calculated. Creating zones can be useful when using plates with larger sized wells, but 96-well plates do not provide enough area per well for zoning to be useful. A mandatory calibration scale must be drawn in the arena settings window. This tells the software how to translate pixels into actual measurements. We use the dynamic subtraction method for our detection settings, which compares the subject color to the background (in our case, the subject is darker than the background). We also incorporate a minimal distance moved threshold that we determined by observing videos of unexposed larvae. This feature filters out the small inconsequential movements of the larvae that are not recognized by a trained observer to be locomotion (i.e., movement from one place to another). If a minimum distance moved threshold is not incorporated, the software will detect slight movements such as head bobbing as locomotor activity. We are only interested in distance moved (e.g., moving from one point to another point in the well) in our results, and we report these distances moved in time bins of 2-min intervals. An important feature of Ethovision software is the ability to calculate a host of endpoints with relative ease. Once acquisition is complete (movements are tracked for each larva) the user can select different endpoints from the list of Dependent Variables provided in the Analysis Profiles section of the software. As mentioned above, we typically focus on distance moved, but there are other movement endpoints, such as velocity, movement (duration for which the animal is changing location, calculated as moving and not moving), acceleration, and acceleration state (high acceleration and low acceleration). Additional variables available include various endpoints for path, direction, and mobility. Each user will need to learn about these endpoints to determine which variable(s) will answer the question(s) being raised in the experiment. When selecting a variable, the software provides a definition of that feature and allows the selection of trial statistics and group statistics for that variable. Once endpoints are selected, the Results section is utilized to calculate the values for dependent variables. Independent variables can be selected from the drop-down menu using the Show/ Hide feature which includes information about start time, video name, and detection settings used, among many others. When the final results with desired endpoints are calculated, we choose to

346

Bridgett N. Hill et al.

export our data to a Microsoft Excel (Office 365) file. In our example, these data are reported as distance moved in centimeters in the time bins selected. Because the results are reported per larva in chronological order per 2 min intervals/time bins (0–2 min interval, 2–4 min interval, etc.), we use an Excel macro to transpose the data so that the results per time bin are displayed across the sheet, utilizing one row for each larva. We find that transposing these data helps with data manipulation (removing any undesired larvae) and analysis (see Subheading 2.11). 2.11 Post-Assay: Statistical Analysis

Deciding whether to label a chemical as having an effect (i.e., a “hit”) is not straightforward. The literature is replete with many distinct approaches to assessing locomotor activity for differences between the treated and control larvae, as well as differences among the various doses of each chemical (e.g., [73, 76–78]). There are probably as many different ways of analyzing zebrafish locomotor activity data as there are laboratories engaged in testing. To date there has not been a rigorous comparison among all the methods, so it is difficult to know the strengths and weaknesses of each. In our laboratory, we have tried many different methods for our analyses. In almost all cases, we are trying to assess if there was an effect of chemical treatment and which concentration groups were different from the control. Two facts that need to be kept in mind when devising a statistical approach: the data are not necessarily normally distributed, and data are being collected on the same larvae over time. One can compare the entire time course of activity, which would mean comparing one curve to another, or one can compare the activity at one portion during the testing. For example, the total activity during the light photoperiod or during the dark photoperiod or both. Using parametric statistics is easiest for comparison because by doing an overall repeated-measure Analysis of Variance (ANOVA) and then determining when interactions are significant, we are automatically correcting for multiple comparisons. Another approach would be to use the ANOVA (parametric) analysis to compare the total activity in the light or dark photoperiods to determine if they are different. If the ANOVA tells us that there is an overall effect of concentration, we usually follow up with a Fisher’s PLSD post hoc test to determine which concentrations are different from control. As we often have many more individuals in our control groups than in our treated groups, and because of the aforementioned lack of normal distribution of the locomotor activity data, we also use a non-parametric approach for assessment of the light or dark activity. Usually, this consists of a Kruskal Wallis test of the control and all concentration groups to determine if there is a significant concentration response. If there is a significant concentration response, we follow up with a Mann-Whitney U test to determine which

Larval Zebrafish Locomotor Activity Assay

347

concentration groups are different from the control. In these comparisons, we also use a correction procedure for multiple comparisons, such as a type of Bonferroni correction. Often, we will analyze the locomotor activity data using both parametric and non-parametric methods to assess data consistency. Recent criticism of the use of stringent statistical cut-offs (e.g., p-value bioformats_package. jar > save. Open ImageJ, click File > Import > image sequence (1).

l

Select folder containing images for treatment and set desired sequence options (2).

l

Z-stack images are collapsed into a maximum projection image: click image > stacks> Z-Project (3).

l

Select “start” slice and “stop” slice and the Projection type (e.g.: sum slice) (4).

l

For fluorescence quantification, click “polygon selections” and delimit the total area of the worm (5), then click analyze> measure (6, 7).

Application of Fluorescence Microscopy and Behavioral Assays to. . .

411

Fig. 2 The step-by-step to process a z-stack image sequence in ImageJ l

To save the image, click file >save as> TIFF> select the destination folder (8).

3.1.3 Neuronal Viability Analysis

The images processed can be analyzed in comparison to a control group. Some endpoints can be examined in vivo as indicatives of neurotoxicity including the neuronal development, punctum, neuronal absence or shrinkage, neuronal gaps, absence of cell bodies, and reduction in intensity of fluorescence [6, 32, 80, 81]. Figure 3, kindly provided by Tao Ke, shows neuronal changes at dopaminergic neurons induced by Methylmercury (5 μM MeHg) exposure. Neurodegeneration can be also classified by range considering some of the endpoints discussed, as used by Schetinger and coworkers [82]. Cholinergic neurodegeneration was ranked from 0 to 3, where 0 meant no degeneration; 1 meant low degeneration; 2 means moderate degeneration, and 3 means high degeneration considering the head, body, and tail of C. elegans.

3.1.4 Neuronal Activity Imaging

Regarding neuronal activity, C. elegans is an ideal model mainly due to its already mapped nervous system for both hermaphrodites and males [83, 84]. The development of imaging methods enables a fantastic expansion in C. elegans neuronal activity field. While in rodents expensive systems must be used to acquire neuronal

412

Omamuyovwi M. Ijomone et al.

Fig. 3 C. elegans dopaminergic neurodegeneration under Methylmercury (5 μM MeHg) exposure. Blue arrows indicate the punctum, red delimited area indicates neuronal absence, and green delimited area indicates the shrinking soma. BZ555 strain worms were focused on the head, L4 larval stage, 60 objective. Kindly provided by Tao Ke

function, the fluorescent GCaMP construct allows C. elegans in vivo four-dimensional imaging of neuronal calcium implicated in both behavior and neuronal activity patterns [ZM9078 strain; hpIs587 (flp-14p::GCaMP6::wCherry + lin-15(+))] [85] and fluorescent voltage reporters can infer the postsynaptic responses [86]. For instance, GCaMP expression is limited to a set of neurons approaches, and the field of neuronal activity imaging is still limited by a lack of tools for robust assignment of all neurons at the same time. 3.2 Behavioral Assays 3.2.1 Basal Slowing Response

Procedure: l NGM plates can be prepared according to the protocol by [87]. These plates are stored at 20
C 16–20 h prior to the assay. l

Two (2) sets of assay plates are prepared accordingly: bacteria spread (seeded) and un-spread (unseeded) plates.

l

Assay plates to be seeded should be freshly spread with bacteria (OP50 or HB101) depending on the species used for cultivating the worms as follows [51, 53]: – Drop approximately a drop of 2 μL bacteria within a circle (bacteria lawn: inner and outer rings of ~1 and 3.5 cm respectively) in the center of the NGM. – Gently spread the bacteria within the boundary of the circle with the bottom of a glass culture tube (sterilize with 70% ethanol and flame). – Place both sets of plates (spread and un-spread) in an incubator overnight at 37
C.

Application of Fluorescence Microscopy and Behavioral Assays to. . .

413

l

Take assay plates out of the incubator and allow to cool at room temperature before use, dry off lids with a Kimwipe.

l

Already cultivated well-fed animals are washed in a buffer (~1.5 mL of M9 or S-basal) after being removed from culture plates before being transferred to assay plates. Washing, to remove bacteria from the worms, is done twice as follows: – First, wash worms off plates using the buffer into Eppendorf tubes. – Spin in a centrifuge 30–60 s at 1000 rpm or allow worms to settle to the bottom of the tube, then pull off buffer with pipette leaving 100–200 μL.

l

Transfer worms (5–10 worms) to the clear zone in the center of the assay plates as follows: – In a drop of the buffer, use a capillary pipette or cut a 200 μL tip to transfer ~5 μL of worms. Ensure there is a minimum of 5 worms on the assay plate.

l

Gently mop up the drop of the buffer used to transfer the worms from the assay plate with a Kimwipe.

l

Allow assay plate sit undisturbed on a bench for about 5 min of adaptation time.

l

Body movement count (movement rate) is counted; thus, – Count the number of bends per 20 s for each worm, i.e., change in body direction moving either forward or backward of the area behind the pharynx. – Frequency of body bends can be recorded manually and digitally for 20–60 s with a data acquisition tracker software [88, 89].

l

The analysis is calculated; thus, Basal slowing response ¼.

[rate of movement in unseeded plates – the rate of movement in seeded plates] / [rate of movement in seeded plates]. Notes (a) Controls adapted for use during the assay should be well fed and similar in size relative to the thickness of the bacteria lawn in each plate.

(b) Due caution should be taken on the purity (freshness) of reagents used as this can affect the assay. (c) Transfer of worms to assay plates should be done one group at a time. This is to allow time to plate the worms and mop up the liquid (buffer) and efficient counting of bends. (d) To prevent worms from crawling off agar of unseeded assay plates, 100 μL of 4 M fructose can be dropped on the edge of

414

Omamuyovwi M. Ijomone et al.

the plate. Tilt gently to spread the fructose all the way around the late. Allow drying for ~10 min before beginning the assay. (e) These experiments are done blindly and repeated at least 3 independent times. (f) Assays can be recorded via digital camera and analyzed later. 3.2.2 Ethanol Avoidance/ Ethanol Preference

Depending on the investigation being carried out, this behavioral procedure can either be tagged an ethanol avoidance or preference assay. For the ethanol avoidance [90]: l

Synchronized young adult worms (~63 h post-synchronization) to an assay plate which has been divided into four quadrants. – Assay plate is prepared thus; Divide 10 cm NGM petri dishes into four quadrants consisting of A and B (ethanol quadrant), C and D (control quadrant), with each pair of quadrants opposite each other. Bore 9 mm holes into quadrants A and B, in which 50 μL of 100% ice-cold ethanol will be poured. Seal the plate with parafilm to allow ethanol equilibration. – Transfer of worms is done by washing them off ethanol plate using S-basal (twice), then once in distilled water.

l

About 100–200 animals are transferred into the assay plate which contains a central marked spot (forms the origin). These worms are allowed to freely move for 30 min, after which scoring is done. A video recording device of choice can be employed for use within this duration.

The preference index is calculated thus: [Number of worms in control quadrants - Number of worms in ethanol quadrants] / Total number of worms tested. For an ethanol preference [58]: l

Worms are pre-exposed to ethanol as follows: synchronized young adult worms (~63 h post-synchronization) are pretreated by incubating for about 4 h in a seeded NGM control or ethanol pre-exposure plate. – Ethanol plate is prepared as follows: half of NGM plate is seeded with OP50 and then allowed to dry for ~2 h. The other half is seeded with 300 mM ice-cold ethanol. Plate is sealed with a parafilm so as to allow adequate diffusion of the ethanol within the agar for another 2 h.

l

Worms are transferred to assay plate as describe above, and procedure similarly performed as above.

Application of Fluorescence Microscopy and Behavioral Assays to. . .

415

The preference index is calculated thus: [Number of worms in ethanol quadrants - Number of worms in control quadrants] / Total number of worms tested. Notes (a) Adequate care should be taken in quantifying the concentration of ethanol used during assay plate preparation.

(b) Avoid overcrowding the assay plate in order to prevent clustering of worms. (c) Ensure the working bench is horizontal and avoid all external sources of disturbances such as wind, centrifuge, and incubator, in order to eliminate vibrations. (d) Do not count clumped worms left in the midline (center point) of the assay plate while calculating the chemotaxis index, as these consist of wash-injured animals. 3.2.3 Plasticity of Chemotaxis by NaCl

Procedure: l Prepare a culture media containing 14 mL of NGM using a 6 cm plate. This volume of NGM is the minimum requirement to get a sufficient spread of bacterial (overnight cultured E. coli) lawn. This should be incubated for a day at room temperature or overnight at 37
C, then stored at room temperature. l

Prepare the four sets of plates as follows [51, 91, 92]: – Conditioning Plate: This medium is made up of a combination of 25 g/L agar, 100 mM NaCl, 1 mM CaCl2, 1 mM MgSO4, 5 mM KH2PO4(pH 6.0). Autoclave, thereafter pour 8 mL into the 9 cm plate (or 9 mL into 10 cm plate). – Mock-conditioning plate: This is prepared with similar constituents as the conditioning plate but with the exemption of NaCl. – NaCl plug plate: The plug medium consists of 17 g/L agar, 100 mM NaCl, and 1 mM CaCl2, 1 mM MgSO4, 5 mM KH2PO4 (pH 6.0). Autoclave, then pour 6 mL of the medium into a 6 cm plate. – Assay plate: This medium is prepared in a similar way as the plug plate but without NaCl. After autoclaving, pour ~3–3.5 mL of the medium into a 6 cm plate.

l

Store all plates at 4
C and use within 2–3 weeks Create a NaCl concentration gradient on the assay plate. This can be done in two ways: – Diffusion method: Use a Pasteur pipette (remove the tip) to excise a chunk of diameter 5 mm from the NaCl plug plate. Place this excised plug on one side of the assay plate for ~19–23 h. This overlapping plug should be discarded before the commencement of the chemotaxis assay [51, 60].

416

Omamuyovwi M. Ijomone et al.

– Spotting method: This is done by directly applying small drops of the chosen concentration of NaCl to a side of the assay plate at an equidistant point from the control region [91, 93]. l

l

l

l

Use a wash buffer to collect well-fed animals from NGM. Wash buffer is made up of 1 mM CaCl2, 1 mM MgSO4, 5 mM KH2PO4 (pH 6.0), and 0.05% Tween 20]. NB-: Wash buffer with or without NaCl can be used for the pre-exposure conditions. Then animals are maintained at 20
C for 15 min in their respective pre-exposure wash solutions. Thereafter, wash once with buffer without NaCl for 60 s at 450  g for 60 s; this is to remove the NaCl from the body of the worms [91, 93]. Wash animals collected from NGM in a centrifuge three times using 1.5 mL tubes for 20 s at ~900  g. Transfer washed animals to the conditioning or mockconditioning plates, respectively. Cover plates with lids, then place in an incubator at 20
C for 4 h.

l

Use wash buffer to collect conditioned and mock-conditioned animals into 1.5 mL tubes.

l

Gently place animals in the center of the assay plate (region having the established NaCl concentrated gradient), approximately 50–300 animals. Use a Kimwipe to mop off excess wash solution.

l

Allow plate with lids to seat undisturbed on the working / experimental bench for 15 min.

l

Put several drops of chloroform on the lid to stop animal movement.

l

Count the total number of animals on each fraction of the assay plate.

l

Calculation of chemotaxis index is done thus [51]; ½X  Y =½X þ Y 

where X ¼ Number of animals on NaCl fraction of the assay plate. Y ¼ Number of animals on the other (control) fraction of the assay plate. Notes (a) Ensure the working bench is horizontal and avoid all external sources of disturbances such as wind, centrifuge, and incubator, in order to eliminate vibrations.

Application of Fluorescence Microscopy and Behavioral Assays to. . .

417

(b) Do not use old seeded nematode growth media (over 2 weeks). (c) Do not count clumped worms left in the midline (center point) of the assay plate while calculating the chemotaxis index, as these consist of wash-injured animals. (d) Presence of food during the assay may alter the results or create biases in the response of worms to the concentration gradient. (e) Before commencement of the assay procedure, ensure animals are well fed and not from a starved agar chunk. (f) Ensure room temperature is kept at 25  1
C during the assay procedure. (g) The concentration of NaCl (between 50 mM and 100 mM) used during the pre-exposure/preconditioning stage should be similar to that used for the assay [51, 59]. 3.2.4 Loopy Foraging and Shrinker Behaviors

Procedures are adapted from this previously described method [63, 94]. Procedure for Loopy Foraging: l At young adult stage (~63 h post synchronization), pick 20 worms unto assay plate. Allow plate to seat on bench for 5 min.

– Assay Plate(prepared a day before assay): Prepare thin seeded NGM plates by dropping ~30 μL OP50 bacterial into 60 mM plates, spread and allow to seat at room temperature for 24 h or overnight at 37
C. l

Observe and count number of worms with “loopy” foraging per 20 worms.

Procedure for Shrinker Behavior: l At young adult stage (~63 h post synchronization), pick 20 worms unto assay plate. Allow plate to seat on bench for 5 min.

– Assay Plate: Prepare thin seeded NGM plates by dropping ~30 μL OP50 bacterial into 60 mM plates, spread and allow to seat at room temperature for 24 h or overnight at 37
C. l

Gently prod or touch body of worm with a pick, and observe “shrinking” action (see fig below).

l

Count number of “shrinkers” per 20 worms.

For examples of phenotypic pattern of loopy foraging and shrinker behaviors see Jorgenson 2005 [63].

418

Omamuyovwi M. Ijomone et al.

Notes l CB156 strain [unc-25(e156) III] worm strains can be used as appropriate controls. These worms which are unc-25 mutants show both loopy head and body shrinking behaviors. l

3.2.5 Locomotion Assay

Same worms can be used for both assays. First observe and count loopy head worms. Then prod worms for body shrinking.

Procedure is modified [94] from the previously described method in [52]. Procedure: l Post synchronized young adult worms (~63 h) are collected by washing thrice in S-basal buffer. l

Transfer ~7–10 worms with S-buffer into unseeded NGM assay plates using a Pasteur pipette. The use of unseeded plates is to eliminate the slowing of movements in bacteria lawn, particularly at the edge of the lawn, to assess the general locomotory rate.

l

Remove excess buffer with a Kimwipe and allow plate seat for 5 min of acclimatization.

l

Count the number of body bends per 3 min. The body bend count is denoted as each time the posterior bulb of the pharynx reaches a maximum turn in the opposing direction from the last count. – The assay can be repeated several times and data collected over different experiments.

l

Data collected is expressed thus and the average calculated; Body bends/3 min.

Notes (a) Care should be taken not to mistake a worm body reversal in backward motion as a body bend. Such reversal occurs when during forward motion, the worm suddenly reverses thereby curving in the same direction it was previously bent/turned in the backward motion.

(b) Prepare one assay session and finish count before preparing another. (c) Worms counted should be removed from the assay plate by gently picking up to prevent double counting. (d) If worms tend to crawl off Agar plate, make fructose (4 M) or glycerol (8 M) rings around unseeded assay plates. Drop 60–100 μL of fructose or glycerol solution on edge of plate,

Application of Fluorescence Microscopy and Behavioral Assays to. . .

419

tilt plate to coat the edge all the way around, and allow to dry for about 10 min. (e) These experiments are done blindly and repeated at least 3 independent times. (f) Assays can be recorded via digital camera and analyzed later. 3.2.6 Pharyngeal Pumping and Thrashing Behavior

Procedure: l Transfer at least 10 worms from synchronized population into freshly prepared NGM assay plate. l

Count the number of times worms pump per minute (ppm) in two ways: – Using a stereomicroscope, you may choose to manually count pumping rate, 10–20 s per min for 10 min [66] OR. – If the pumping rate seems abnormal, one can decide to digitally count the pumping rate with a digital camera (at least 10 frames per sec). It is preferred to transfer a single worm into the NGM plate with a bacteria lawn of 5 mm. – Counts could also be done via another technique called the high power differential interference contrast [66].

l

Data is calculated thus; the rate of pharyngeal pumping ¼ Total number of pumps/Total time For the thrashing assay, plates can be prepared in two ways: – Put 10 μL of dH2O in a shallow transparent Petri plate, then transfer worms with a Pasteur pipette into this assay plate [50] OR. – Place the worm in the middle of a droplet of the isotonic solution in a standard NGM. Counting is done as the worm tries to reach a solid medium [65]. – Thrashing assay is monitored via videotaping and assessed using any available recommended assay software [50].

Notes (a) Make use of well-fed animals, as pharyngeal pumping is decreased in normal animals that are off food.

(b) Replay recorded videos of pumping assays at ½ to 1/3 of the original speed to aid accurate count. (c) Make use of standard NGM that are well seeded with bacteria to avoid irregular increased motility of hungry worms. Such would make pump count difficult. (d) L1 worms require a magnification of X100 for accurate viewing.

420

Omamuyovwi M. Ijomone et al.

(e) These experiments are done blindly and repeated at least 3 independent times. (f) Assays can be recorded via digital camera and analyzed later. 3.3 Behavioral Assay Under Control of Optogenetics

For more detailed description see [76, 79]. Procedure: l Prepare NGM using 12 mL of agar, then store at room temperature for 48 h before use. l

Seed NGM with a 50 μL stock mixture of bacteria and ATR in a ratio 49.4 μL:0.6 μL, onto the center of the NGM forming a square (4  4 cm area).

l

Store seeded plates in the dark (using a foil or kept in a drawer) or under red light to allow bacteria growth (~24 h).

l

Age synchronize worms (use 3–4-day old young adult worms). This can be achieved by adapting the use of gravid strains that are allowed a 4 h egg-laying period within the plates. These are later removed to have 40 80 age synchronized worms. – To achieve a strain of intense light response, worms expressing extrachromosomal array can be transferred into fresh ATR plates.

l

Transfer worms into seeded plates by picking using a Pasteur pipette or pipette with M9 buffer.

l

Seal worms with parafilm. Store in a controlled dark environment.

l

Attach the blue light filter over the lens of the Multi-Worm Tracker.

l

The custom LED light (supplied by the DC power) should then be connected to the stimulus relay of the multi-worm tracker.

l

Launch multi-worm tracker software, arrange the light stimulus parameter.

l

Begin mounting of each plate on the tracker platform, focus the LED light accordingly at the center of the ring on the plates.

l

Analyze data using offline java-command-line Chorography software; use custom scripts to re-arrange output files as desired, investigating neuronal activities and behavioral changes.

Notes (a) Prepare extra seeded plates in case of contamination or uneven distribution of agar spread.

(b) Ensure to minimize the ATR exposure to light to avoid the photosensitizer losing its gating function on the opsins.

Application of Fluorescence Microscopy and Behavioral Assays to. . .

421

(c) The seeded plates should be used within 72 h to avoid excess bacteria growth. (d) Reset the worm tracker each time the plate is changed.

4

Conclusions The set of connections in neural systems called the connectome form the communication in the brain. Detailed understanding of these connections, i.e., mapping high-resolution connectivity, is an essential first step in elucidating how the nervous system processes information and generates behavior [95]. C. elegans is an ideal model for studying neuronal activity due to its small, stereotyped yet relatively complex nervous system [84]. Furthermore, its optical transparency, homologous genomic constitution with humans and availability is an added advantage. Knowledge about nervous system connectivity is critical to understanding how it functions. Cook et al. (2019) described in their studies the concepts of both adult sexual forms of the nematode C. elegans. Serial transmission electron microscopy and some previously related data were used to perform a reconstruction of circuits for the male head, including mainly the nerve ring and retro-vesicular globule [96]. The evaluation of whole-animal connectomes from sensory input to end-organ output across the worm showed a considerable number of connections’ differences between the male and hermaphrodites [96]. The connectivity profile can indicate how neurons work, whether by sensory perception or hormone secretion. A thorough understanding of the interconnections of various neurotransmitter systems of this nematode is imperative to understand certain behavioral outputs among species. Although varying protocols could be modified, the use of standard methods is essential to ensure reproducibility of results.

Acknowledgments OMI acknowledges the 2019 Young IBRO Regions Connecting Awards. MA is supported by National Institute of Health (NIH), USA grants, NIEHS R01 10563, NIEHS R01 07331, and NIEHS R01 020852. We acknowledge Tao Ke of the Albert Einstein College of Medicine for images of C. elegans dopaminergic neurodegeneration under MeHg exposure. References 1. Consortium CeS (1998) Genome sequence of the nematode C-elegans: a platform for investigating biology. Science 282

(5396):2012–2018. https://doi.org/10. 1126/science.282.5396.2012

422

Omamuyovwi M. Ijomone et al.

2. Antoshechkin I, Sternberg PW (2007) The versatile worm: genetic and genomic resources for Caenorhabditis elegans research. Nat Rev Genet 8(7):518–532. https://doi.org/10. 1038/nrg2105 3. Chalfie M, Tu Y, Euskirchen G, Ward WW, Prasher DC (1994) Green fluorescent protein as a marker for gene-expression. Science 263 (5148):802–805. https://doi.org/10.1126/ science.8303295 4. Boulin T, Pocock R, Hobert O (2006) A novel Eph receptor-interacting IgSF protein provides C. elegans motoneurons with midline guidepost function. Curr Biol 16(19):1871–1883. https://doi.org/10.1016/j.cub.2006.08.056 5. Feinberg EH, VanHoven MK, Bendesky A, Wang G, Fetter RD, Shen K, Bargmannl CI (2008) GFP reconstitution across synaptic partners (GRASP) defines cell contacts and synapses in living nervous systems. Neuron 57 (3):353–363. https://doi.org/10.1016/j.neu ron.2007.11.030 6. Benedetto A, Au C, Avila DS, Milatovic D, Aschner M (2010) Extracellular dopamine potentiates Mn-induced oxidative stress, lifespan reduction, and dopaminergic neurodegeneration in a BLI-3-dependent manner in Caenorhabditis elegans. PLoS Genet 6(8): e1001084. https://doi.org/10.1371/journal. pgen.1001084 7. Gubert P, Puntel B, Lehmen T, Fessel JP, Cheng P, Bornhorst J, Trindade LS, Avila DS, Aschner M, Soares FAA (2018) Metabolic effects of manganese in the nematode Caenorhabditis elegans through DAergic pathway and transcription factors activation. Neurotoxicology 67:65–72. https://doi.org/10.1016/j. neuro.2018.04.008 8. DiLoreto EM, Chute CD, Bryce S, Srinivasan J (2019) Novel technological advances in functional connectomics in C. elegans. J Dev Biol 7 (2). https://doi.org/10.3390/jdb7020008 9. Marsh EK, May RC (2012) Caenorhabditis elegans, a model organism for investigating immunity. Appl Environ Microbiol 78 (7):2075–2081 10. Johnson TE (2003) Advantages and disadvantages of Caenorhabditis elegans for aging research. Exp Gerontol 38(11–12):1329–1332 11. Gottschling D-C, Do¨ring F (2019) Is C. elegans a suitable model for nutritional science? Genes Nutr 14(1):1–4 12. Tissenbaum HA (2015) Using C. elegans for aging research. Invertebr Reprod Dev 59 (sup1):59–63 13. Leung MC, Williams PL, Benedetto A, Au C, Helmcke KJ, Aschner M, Meyer JN (2008)

Caenorhabditis elegans: an emerging model in biomedical and environmental toxicology. Toxicol Sci 106(1):5–28 14. Chen X, Barclay JW, Burgoyne RD, Morgan A (2015) Using C. elegans to discover therapeutic compounds for ageing-associated neurodegenerative diseases. Chem Cent J 9(1):65 15. Alexander AG, Marfil V, Li C (2014) Use of Caenorhabditis elegans as a model to study Alzheimer’s disease and other neurodegenerative diseases. Front Genet 5:279 16. Hobert O (2013) The neuronal genome of Caenorhabditis elegans. WormBook:1–106. https://doi.org/10.1895/wormbook.1. 161.1 17. Loer CMR, Rand JB (2016) The evidence for classical neurotransmitters in C. elegans neurons (updated online review/database in WormAtlas 2.0; original in 2010). WormAtlas 18. Rand JB (2007) Acetylcholine. WormBook:1–21. https://doi.org/10.1895/ wormbook.1.131.1 19. Chase DL, Koelle MR (2007) Biogenic amine neurotransmitters in C. elegans. WormBook:1–15. https://doi.org/10.1895/ wormbook.1.132.1 20. Vidal-Gadea AG, Davis S, Becker L, PierceShimomura JT (2012) Coordination of behavioral hierarchies during environmental transitions in Caenorhabditis elegans. Worm 1 (1):5–11. https://doi.org/10.4161/worm. 19148 21. Tang B, Tong P, Xue KS, Williams PL, Wang JS, Tang L (2019) High-throughput assessment of toxic effects of metal mixtures of cadmium(cd), lead(Pb), and manganese(Mn) in nematode Caenorhabditis elegans. Chemosphere 234:232–241. https://doi.org/10. 1016/j.chemosphere.2019.05.271 22. Baesler J, Kopp JF, Pohl G, Aschner M, Haase H, Schwerdtle T, Bornhorst J (2019) Zn homeostasis in genetic models of Parkinson’s disease in Caenorhabditis elegans. J Trace Elem Med Biol 55:44–49. https://doi.org/10. 1016/j.jtemb.2019.05.005 23. Xing X, Guo Y, Wang D (2009) Using the larvae nematode Caenorhabditis elegans to evaluate neurobehavioral toxicity to metallic salts. Ecotoxicol Environ Saf 72 (7):1819–1823. https://doi.org/10.1016/j. ecoenv.2009.06.006 24. Melo-Thomas L, Engelhardt KA, Thomas U, Hoehl D, Thomas S, Wohr M, Werner B, Bremmer F, Schwarting RKW (2017) A wireless, bidirectional Interface for in vivo recording and stimulation of neural activity in freely

Application of Fluorescence Microscopy and Behavioral Assays to. . . behaving rats. J Vis Exp 129. https://doi.org/ 10.3791/56299 25. Mann T, Kurth J, Hawlitschka A, Stenzel J, Lindner T, Polei S, Hohn A, Krause BJ, Wree A (2018) [(18)F]fallypride-PET/CT analysis of the dopamine D(2)/D(3) receptor in the hemiparkinsonian rat brain following intrastriatal botulinum neurotoxin a injection. Molecules 23(3). https://doi.org/10.3390/ molecules23030587 26. Sands B, Burnaevskiy N, Yun SR, Crane MM, Kaeberlein M, Mendenhall A (2018) A toolkit for DNA assembly, genome engineering and multicolor imaging for C. elegans. Transl Med Aging 2:1–10. https://doi.org/10.1016/j. tma.2018.01.001 27. Green RA, Audhya A, Pozniakovsky A, Dammermann A, Pemble H, Monen J, Portier N, Hyman A, Desai A, Oegema K (2008) Expression and imaging of fluorescent proteins in the C-elegans gonad and early embryo. Method Cell Biol 85:179. https:// doi.org/10.1016/S0091-679x(08)85009-1 28. Heppert JK, Dickinson D, Pani AM, Higgins CD, Goldstein B (2016) Comparative assessment of fluorescent proteins for in vivo imaging in an animal model system. Mol Biol Cell 27. ://WOS:000396046900067 29. Molino JVD, de Carvalho JCM, Mayfield S (2018) Evaluation of secretion reporters to microalgae biotechnology: blue to red fluorescent proteins. Algal Res 31:252–261. https:// doi.org/10.1016/j.algal.2018.02.018 30. Kobayashi J, Shidara H, Morisawa Y, Kawakami M, Tanahashi Y, Hotta K, Oka K (2013) A method for selective ablation of neurons in C. elegans using the phototoxic fluorescent protein, KillerRed. Neurosci Lett 548:261–264. https://doi.org/10.1016/j. neulet.2013.05.053 31. Flames N, Hobert O (2009) Gene regulatory logic of dopamine neuron differentiation. Nature 458(7240):885–889. https://doi. org/10.1038/nature07929 32. Nass R, Hall DH, Miller DM 3rd, Blakely RD (2002) Neurotoxin-induced degeneration of dopamine neurons in Caenorhabditis elegans. Proc Natl Acad Sci U S A 99(5):3264–3269. https://doi.org/10.1073/pnas.042497999 33. Dauer W, Przedborski S (2003) Parkinson’s disease: mechanisms and models. Neuron 39 (6):889–909 34. Harrington AJ, Yacoubian TA, Slone SR, Caldwell KA, Caldwell GA (2012) Functional analysis of VPS41-mediated neuroprotection in Caenorhabditis elegans and mammalian

423

models of Parkinson’s disease. J Neurosci 32 (6):2142–2153 35. Serrano-Saiz E, Poole RJ, Felton T, Zhang F, De La Cruz ED, Hobert O (2013) Modular control of glutamatergic neuronal identity in C. elegans by distinct homeodomain proteins. Cell 155(3):659–673. https://doi.org/10. 1016/j.cell.2013.09.052 36. Lieke T, Steinberg CE, Ju J, Saul N (2015) Natural marine and synthetic xenobiotics get on nematode’s nerves: neuro-stimulating and neurotoxic findings in Caenorhabditis elegans. Mar Drugs 13(5):2785–2812. https://doi. org/10.3390/md13052785 37. Soares FA, Fagundez DA, Avila DS (2017) Neurodegeneration induced by metals in Caenorhabditis elegans. Adv Neurobiol 18:355–383. https://doi.org/10.1007/9783-319-60189-2_18 38. Jacques MT, Bornhorst J, Soares MV, Schwerdtle T, Garcia S, Avila DS (2019) Reprotoxicity of glyphosate-based formulation in Caenorhabditis elegans is not due to the active ingredient only. Environ Pollut 252 (Pt B):1854–1862. https://doi.org/10. 1016/j.envpol.2019.06.099 39. Zhang RQ, Hong Y, Xiao JS (2013) Separation and determination of pyrrolidinium ionic liquid cations by ion chromatography with direct conductivity detection. Chin Chem Lett 24 (6):503–505 40. Rohn I, Raschke S, Aschner M, Tuck S, Kuehnelt D, Kipp A, Schwerdtle T, Bornhorst J (2019) Treatment of Caenorhabditis elegans with small selenium species enhances antioxidant defense systems. Mol Nutr Food Res 63 (9). https://doi.org/10.1002/mnfr. 201801304 41. Gubert P, Puntel B, Lehmen T, Bornhorst J, Avila DS, Aschner M, Soares FAA (2016) Reversible reprotoxic effects of manganese through DAF-16 transcription factor activation and vitellogenin downregulation in Caenorhabditis elegans. Life Sci 151:218–223. https://doi.org/10.1016/j.lfs.2016.03.016 42. Kim W, Underwood RS, Greenwald I, Shaye DD (2018) OrthoList 2: a new comparative genomic analysis of human and Caenorhabditis elegans genes. Genetics 210(2):445–461 43. Corsi AK, Wightman B, Chalfie M (2015) A transparent window into biology: a primer on Caenorhabditis elegans. Genetics 200 (2):387–407 44. Silverman GA, Luke CJ, Bhatia SR, Long OS, Vetica AC, Perlmutter DH, Pak SC (2009) Modeling molecular and cellular aspects of

424

Omamuyovwi M. Ijomone et al.

human disease using the nematode Caenorhabditis elegans. Pediatr Res 65(1):10 45. Metaxakis A, Petratou D, Tavernarakis N (2018) Multimodal sensory processing in Caenorhabditis elegans. Open Biol 8(6):180049 ´ lvarez R, Otaduy MA, 46. Mujika A, Lesˇkovsky´ P, A Epelde G (2017) Modeling behavioral experiment interaction and environmental stimuli for a synthetic C. elegans. Front Neuroinform 11:71 47. Maulik M, Mitra S, Bult-Ito A, Taylor BE, Vayndorf EM (2017) Behavioral phenotyping and pathological indicators of Parkinson’s disease in C. elegans models. Frontiers in genetics 8:77 48. Rotroff DM, Joubert BR, Marvel SW, Ha˚berg SE, Wu MC, Nilsen RM, Ueland PM, Nystad W, London SJ, Motsinger-Reif A (2016) Maternal smoking impacts key biological pathways in newborns through epigenetic modification in utero. BMC Genomics 17(1):976 49. Raley-Susman KM, Chou E, Lemoine H (2018) Use of the model organism Caenorhabditis elegans to elucidate neurotoxic and behavioral effects of commercial fungicides. Neurotoxins 37 50. Chen P, Martinez-Finley EJ, Bornhorst J, Chakraborty S, Aschner M (2013) Metalinduced neurodegeneration in C. elegans. Front Aging Neurosc 5:18 51. Ijomone OM, Miah MR, Peres TV, Nwoha PU, Aschner M (2016) Null allele mutants of trt-1, the catalytic subunit of telomerase in Caenorhabditis elegans, are less sensitive to Mn-induced toxicity and DAergic degeneration. Neurotoxicology 57:54–60. https://doi. org/10.1016/j.neuro.2016.08.016 52. Hart AC (2006) Behavior (July 3, 2006), WormBook, ed. the C. elegans research community, WormBook. https://doi.org/10. 1895/wormbook.1.87.1 53. Sawin ER, Ranganathan R, Horvitz HR (2000) C. elegans locomotory rate is modulated by the environment through a dopaminergic pathway and by experience through a serotonergic pathway. Neuron 26(3):619–631 54. Aschner M, Chen P, Martinez-Finley EJ, Bornhorst J, Chakraborty S (2013) Metalinduced neurodegeneration in C. elegans. Frontiers in aging neuroscience 5:18 55. Kim H, Calatayud C, Guha S, Ferna´ndezCarasa I, Berkowitz L, Carballo-Carbajal I, Ezquerra M, Ferna´ndez-Santiago R, Kapahi P, ´ (2018) The small GTPase RAC1/ Raya A CED-10 is essential in maintaining dopaminergic neuron function and survival against

α-Synuclein-induced toxicity. Mol Neurobiol 55(9):7533–7552 56. Engleman EA, Katner SN, Neal-Beliveau BS (2016) Caenorhabditis elegans as a model to study the molecular and genetic mechanisms of drug addiction. In: Progress in molecular biology and translational science, vol 137. Elsevier, pp 229–252 57. Zhang Y, Lu H, Bargmann CI (2005) Pathogenic bacteria induce aversive olfactory learning in Caenorhabditis elegans. Nature 438(7065):179–184 58. Lee J, Jee C, McIntire SL (2009) Ethanol preference in C. elegans. Genes Brain Behav 8 (6):578–585 59. Hukema RK, Rademakers S, Jansen G (2008) Gustatory plasticity in C. elegans involves integration of negative cues and NaCl taste mediated by serotonin, dopamine, and glutamate. Learn Mem 15(11):829–836 60. Saeki S, Yamamoto M, Iino Y (2001) Plasticity of chemotaxis revealed by paired presentation of a chemoattractant and starvation in the nematode Caenorhabditis elegans. J Exp Biol 204 (10):1757–1764 61. Oda S, Tomioka M, Iino Y (2011) Neuronal plasticity regulated by the insulin-like signaling pathway underlies salt chemotaxis learning in Caenorhabditis elegans. J Neurophysiol 106 (1):301–308 62. Zhou X, Bessereau J-L (2019) Molecular architecture of genetically-tractable GABA synapses in C. elegans. Front Mol Neurosci 12:304 63. Jorgensen EM (2005) GABA (August 31, 2005), WormBook, ed. The C. elegans research community. WormBook. https://doi. org/10.1895/wormbook.1.14.1 64. Gjorgjieva J, Biron D, Haspel G (2014) Neurobiology of Caenorhabditis elegans locomotion: where do we stand? Bioscience 64 (6):476–486 65. Trojanowski NF, Raizen DM, Fang-Yen C (2016) Pharyngeal pumping in Caenorhabditis elegans depends on tonic and phasic signaling from the nervous system. Sci Rep 6:22940 66. Raizen D, Song B-M, Trojanowski N, You Y-J (2018) Methods for measuring pharyngeal behaviors. In: WormBook: the online review of C. elegans biology [internet]. WormBook 67. Kompoliti K, Verhagen L (2010) Encyclopedia of movement disorders, vol 1. Academic Press 68. Shi Y, Toga AW (2017) Connectome imaging for mapping human brain pathways. Mol Psychiatry 22(9):1230 69. Azulay A, Itskovits E, Zaslaver A (2016) The C. elegans connectome consists of

Application of Fluorescence Microscopy and Behavioral Assays to. . . homogenous circuits with defined functional roles. PLoS Comput Biol 12(9):e1005021 70. Cook SJ, Jarrell TA, Brittin CA, Wang Y, Bloniarz AE, Yakovlev MA, Nguyen KC, Tang LT-H, Bayer EA, Duerr JS (2019) Wholeanimal connectomes of both Caenorhabditis elegans sexes. Nature 571(7763):63–71 71. Lim DH, LeDue J (2017) What is optogenetics and how can we use it to discover more about the brain? Front Young Minds 5 72. Fang-Yen C, Alkema MJ (1677) Samuel AD (2015) illuminating neural circuits and behaviour in Caenorhabditis elegans with optogenetics. Philos Trans R Soc B 370:20140212 73. Guru A, Post RJ, Ho Y-Y, Warden MR (2015) Making sense of optogenetics. Int J Neuropsychopharmacol 18(11):pyv079 74. Schild LC, Glauser DA (2015) Dual color neural activation and behavior control with chrimson and CoChR in Caenorhabditis elegans. Genetics 200(4):1029–1034 75. Husson SJ, Gottschalk A, Leifer AM (2013) Optogenetic manipulation of neural activity in C. elegans: from synapse to circuits and behaviour. Biol Cell 105(6):235–250 76. Yu AJ, McDiarmid TA, Ardiel EL, Rankin CH (2019) High-throughput analysis of behavior under the control of Optogenetics in Caenorhabditis elegans. Curr Protoc Neurosci 86(1): e57 77. Rost BR, Schneider-Warme F, Schmitz D, Hegemann P (2017) Optogenetic tools for subcellular applications in neuroscience. Neuron 96(3):572–603 78. Yu J, Chen K, Lucero RV, Ambrosi CM, Entcheva E (2015) Cardiac optogenetics: enhancement by all-trans-retinal. Sci Rep 5:16542 79. Pokala N, Glater EE (2018) Using optogenetics to understand neuronal mechanisms underlying behavior in C. elegans. J Undergrad Neurosci Educ 16(2):A152 80. Martinez-Finley EJ, Avila DS, Chakraborty S, Aschner M (2011) Insights from Caenorhabditis elegans on the role of metals in neurodegenerative diseases. Metallomics 3 (3):271–279. https://doi.org/10.1039/ c0mt00064g 81. Qu M, Kong Y, Yuana Y, Wang D (2019) Neuronal damage induced by nanopolystyrene particles in nematode Caenorhabditis elegans. Environ Sci Nano 6:2591–2601. https://doi. org/10.1039/C9EN00473D 82. Schetinger MRC, Peres TV, Arantes LP, Carvalho F, Dressler V, Heidrich G, Bowman AB, Aschner M (2019) Combined exposure to methylmercury and manganese during L1

425

larval stage causes motor dysfunction, cholinergic and monoaminergic up-regulation and oxidative stress in L4 Caenorhabditis elegans. Toxicology 411:154–162. https://doi.org/ 10.1016/j.tox.2018.10.006 83. White JG, Southgate E, Thomson JN, Brenner S (1986) The structure of the nervous system of the nematode Caenorhabditis elegans. Philos Trans R Soc Lond Ser B Biol Sci 314 (1165):1–340. https://doi.org/10.1098/ rstb.1986.0056 84. Jarrell TA, Wang Y, Bloniarz AE, Brittin CA, Xu M, Thomson JN, Albertson DG, Hall DH, Emmons SW (2012) The connectome of a decision-making neural network. Science 337 (6093):437–444. https://doi.org/10.1126/ science.1221762 85. Chen TW, Wardill TJ, Sun Y, Pulver SR, Renninger SL, Baohan A, Schreiter ER, Kerr RA, Orger MB, Jayaraman V, Looger LL, Svoboda K, Kim DS (2013) Ultrasensitive fluorescent proteins for imaging neuronal activity. Nature 499(7458):295–300. https://doi.org/ 10.1038/nature12354 86. Piatkevich KD, Jung EE, Straub C, Linghu C, Park D, Suk HJ, Hochbaum DR, Goodwin D, Pnevmatikakis E, Pak N, Kawashima T, Yang CT, Rhoades JL, Shemesh O, Asano S, Yoon YG, Freifeld L, Saulnier JL, Riegler C, Engert F, Hughes T, Drobizhev M, Szabo B, Ahrens MB, Flavell SW, Sabatini BL, Boyden ES (2018) A robotic multidimensional directed evolution approach applied to fluorescent voltage reporters. Nat Chem Biol 14(4):352–360. https://doi.org/10.1038/s41589-018-00049 87. Chaudhuri J, Parihar M, Pires-daSilva A (2011) An introduction to worm lab: from culturing worms to mutagenesis. J Vis Exp 47:e2293 88. Rivard L, Srinivasan J, Stone A, Ochoa S, Sternberg PW, Loer CM (2010) A comparison of experience-dependent locomotory behaviors and biogenic amine neurons in nematode relatives of Caenorhabditis elegans. BMC Neurosci 11(1):22 89. Swierczek NA, Giles AC, Rankin CH, Kerr RA (2011) High-throughput behavioral analysis in C. elegans. Nat Methods 8(7):592 90. Cooper JF, Dues DJ, Spielbauer KK, Machiela E, Senchuk MM, Van Raamsdonk JM (2015) Delaying aging is neuroprotective in Parkinson’s disease: a genetic analysis in C. elegans models. NPJ Parkinson’s Dis 1 (1):1–12 91. Matsuura T, Urushihata T (2015) Chronic nicotine exposure augments gustatory plasticity in Caenorhabditis elegans: involvement of

426

Omamuyovwi M. Ijomone et al.

dopamine signaling. Biosci Biotechnol Biochem 79(3):462–469 92. Urushihata T, Wakabayashi T, Osato S, Yamashita T, Matsuura T (2016) Short-term nicotine exposure induces long-lasting modulation of gustatory plasticity in Caenorhabditis elegans. Biochem Biophys Rep 8:41–47 93. Urushihata T, Takuwa H, Higuchi Y, Sakata K, Wakabayashi T, Nishino A, Matsuura T (2016) Inhibitory effects of caffeine on gustatory plasticity in the nematode Caenorhabditis elegans. Biosci Biotechnol Biochem 80 (10):1990–1994 94. Ijomone OM, Miah MR, Akingbade GT, Bucinca H, Aschner M (2020) Nickel-induced developmental neurotoxicity in C. elegans includes cholinergic, dopaminergic and

GABAergic degeneration, altered behaviour, and increased SKN-1 activity. Neurotox Res 37:1010–1028. https://doi.org/10.1007/ s12640-020-00175-3 95. Bhattacharya A, Aghayeva U, Berghoff EG, Hobert O (2019) Plasticity of the electrical connectome of C. elegans. Cell 176 (5):1174–1189. e1116. https://doi.org/10. 1016/j.cell.2018.12.024 96. Cook SJ, Jarrell TA, Brittin CA, Wang Y, Bloniarz AE, Yakovlev MA, Nguyen KCQ, Tang LT, Bayer EA, Duerr JS, Bulow HE, Hobert O, Hall DH, Emmons SW (2019) Whole-animal connectomes of both Caenorhabditis elegans sexes. Nature 571(7763):63–71. https://doi. org/10.1038/s41586-019-1352-7

Part VI In Vitro Methods

Chapter 19 Effects of Neurotoxic or Pro-regenerative Agents on Motor and Sensory Neurite Outgrowth in Spinal Cord Organotypic Slices and DRG Explants in Culture Sara Bolı´var, Ilary Allodi, Mireia Herrando-Grabulosa, and Esther Udina Abstract Classically, primary sensory neuron cultures obtained from the DRG have been used as a model to evaluate neurite growth in vitro. Primary sensory neurons are easily cultured, either dissociated or from explants, from embryonic to adult ages. In contrast, culture of motoneurons is much more complex and limited to the embryonic ones or to postnatal organotypic cultures by using membrane culture inserts. Here we describe a protocol of an easy in vitro assay to culture postnatal rodent spinal cord organotypic slices and DRG explants in 3D collagen matrices that are permissive for neuritogenesis. The main aim of this in vitro assay is to have a similar setting for both types of neurons that allows the measurement and comparison of positive or adverse events on neurite growth of motor and sensory neurons. The matrix can also be modified by adding trophic or tropic factors, cells, or other agents. Immunohistochemistry of the explants and the slices is needed to specifically label myelinated fibers and fairly compare the growth of myelinated primary sensory neurons and motoneurons, as well as neuronal survival. Key words Motoneuron, Dorsal root ganglia, Spinal cord, Primary sensory neurons, Neurite growth, Collagen matrix, Explant, Organotypic, Postnatal, Mice, Rat

1

Introduction Spinal motoneurons and primary sensory neurons of the dorsal root ganglia (DRG) are the main contributors of the axons that form peripheral nerves, together with autonomic axons. In contrast to the central nervous system, the peripheral nervous system lacks the blood-brain barrier, which usually protects the brain and the spinal cord from external insults. This fact explains why peripheral axons are more prone to suffer after exposure to neurotoxic drugs. A paradigmatic example is the peripheral neuropathy induced by neurotoxic chemotherapy drugs, extensively used to treat different types of cancers [1]. Although these axonopathic agents would affect both motor and sensory axons, DRG neurons are more commonly affected since their somas are more exposed. Therefore,

Jordi Llorens and Marta Barenys (eds.), Experimental Neurotoxicology Methods, Neuromethods, vol. 172, https://doi.org/10.1007/978-1-0716-1637-6_19, © Springer Science+Business Media, LLC, part of Springer Nature 2021

429

430

Sara Bolı´var et al.

in vitro settings to study peripheral neurotoxicity have mainly focused on sensory neuronal cultures [2]. Similarly, since rodent primary sensory neurons are easily cultured, from embryonic to adult, in vitro studies regarding neurite outgrowth have also been mainly focused on DRG cultures, either DRG explants or dissociated primary sensory neurons [3]. In contrast, neurotoxicity and neurodegenerative conditions in motoneurons are evaluated using spinal cord organotypic slices [4, 5]. The culture of adult motoneurons is quite complex and few papers reported success [6, 7]. Therefore, classical spinal cord explants are performed in postnatal rats by using permeable membrane culture inserts [8, 9]. These inserts allow long-term survival of spinal cord slices but suppose some constraints regarding neurite growth evaluation. Since sensory and motor neuron cultures are classically performed separately and using different settings, an in vitro setting that allows a fair comparison between both of them would be highly advantageous. Here we describe an easy in vitro assay to culture postnatal spinal cord organotypic slices and DRG explants in collagen matrices [10] that are permissive for neuritogenesis. The aim of this in vitro model is to have a similar setting for both types of neurons that facilitates the comparative measurement of positive or adverse events on neurite growth of motor and sensory neurons [11]. Although usually this type of culture has been performed using postnatal rats, it can be adapted to postnatal mice. This is of great interest since transgenic mice have become a valuable tool for research.

2

Materials A table with the reference of the bioreagents and surgical tools used in this protocol is included (Table 1).

2.1 Spinal Cord Organotypic Slices and DRG Explants Cultures

l

Light microscope (for dissection). A portable dissection microscope that can be placed into the cell culture hood is recommended for fine cleaning of the tissue, whereas extraction of the primary sample can be performed either in a portable or a fixed one.

l

Basic surgical tools (scalp, fine forceps, laminectomy forceps, spring scissors. . .).

l

Gey’s balanced salt solution enriched with 6 mg/mL sterile D(+)-glucose. Make fresh as required and store at 4
C.

DRG and Spinal Cord Organotypic Cultures on Collagen Matrices

431

Table 1 Bioreagents and surgical tools used in the protocol, with recommended vendor and reference number Product

Reference

Vendor

B27 supplement (50)

17504044

ThermoFisher (Gibco)

Gey’s balanced salt solution

G9779

Sigma

L-Glutamine

G7513

Sigma

Minimum essential medium (MEM) 10

11430030

ThermoFisher (Gibco)

Neurobasal medium

21103049

ThermoFisher (Gibco)

Nylon net filter

HNWP04700

Millipore

Penicillin-streptomycin

P0781

Sigma

Poly-D-lysine hydrobromide

P6407

Sigma

Rat tail type I collagen

354236

Corning

Sodium bicarbonate 7.5%

25080-094

ThermoFisher (Gibco)

Sterile D-(+)-glucose

G7021

Sigma

Trizma hydrochloride (HCl)

T5941

Sigma

Trizma base

T1503

Sigma

Tris buffer (TB; 0.05 M pH 70 4)

Trizma HCl (6.06 g), Trizma Base (1.39 g), distilled H2O (1 L)

Tris buffer saline (TBS; 0.05 M pH 70 4)

TB10  (100 mL), NaCl (8 g), distilled-H2O (900 mL)

solution

Surgical tools Small spring scissors

FST 15003-08

Fine science tools

Spring scissors

FST 15025-10

Fine science tools

Dumont 2 (laminectomy)

FST 11223-20

Fine science tools

Dumont 5 (fine forceps)

FST 11254-20

Fine science tools

l

Ice pack to keep the samples cold. A homemade ice pack can be easily made by using glass Petry dishes of 60 mm diameter filled with gel from commercial ice packs, sealed and kept on the freezer before use. Its flat surface and small size are ideal as a base to manipulate the samples under the dissection microscope.

l

Cell culture hood with laminar flux.

l

McIlwain Tissue Chopper.

l

Petri dishes or 24 multiwell plates.

432

Sara Bolı´var et al. l

Round coverslips.

l

Poly-D-lysine (10 μg/mL).

l

2.2 Immunohistochemistry

2.3

3

Quantification

Rat tail type I collagen solution (3.4 mg/mL); store at 4
C and keep cold during use.

l

10 Minimum essential medium (MEM).

l

7.5% Sodium bicarbonate solution (can be stored at 4
C).

l

Culture incubator set at 37
C and 5% CO2.

l

Neurobasal medium (NB), supplemented with B27 (I1), glutamine (20 mM), glucose (6 mg/mL) and with or without penicillin/streptomycin (1: 100 U/mL and 0.1 mg/mL, respectively). Make fresh for each culture and store at 4
C.

l

Nylon hydrophilic membrane filter.

l

Paraformaldehyde (PFA) 4% pH 7,4, store at 4
C but warm at 37
C before use.

l

Tris Buffer (TB), Tris-buffered saline (TBS), Tris-buffered saline with 0.3% Triton (TBSL).

l

Citrate buffer (10 mM, pH 6.1).

l

Methanol (50%, 70%, and 100%).

l

Ethanol (70%, 96%, and 100%).

l

Mounting media (for example, glycerol supplemented with 10% Mowiol and 0.6% DABCO (Sigma), or Fluoromount (SouthernBiotech)).

l

Inverted light microscope (to evaluate fresh cultures).

l

Epifluorescence microscope and/or confocal microscope (to evaluate immunocytochemistry processed slices).

l

Image J software (NIH, available at http://rsb.info.nih.gov/ij/ ).

Methods

3.1 Matrix Preparation 3.1.1 Preliminary Preparations

The explants are placed on top of a collagen matrix that needs to be prepared before harvesting the samples since its gelation needs around 2 h. Coating of the coverslips placed on the culture dishes is recommended to facilitate attachment of the collagen matrices, and therefore it is better to prepare it the day before.

DRG and Spinal Cord Organotypic Cultures on Collagen Matrices

433

3.1.2 Preparation of the Matrices

Coverslips are placed in Petri dishes or multiwell plates (three coverslips in a 60 mm petri dish or single ones in a 24-well plate, see Note1) and coated with poly-D-lysine (PDL, 10 μg/mL) overnight. Three washes with sterile distilled water are performed afterward. Coverslips have to be completely dry to avoid that water droplets dilute the collagen matrix. The collagen matrices can be prepared in 1.5 mL Eppendorf (500 μL each) and kept on ice to avoid gelation until it is used. 450 μL of rat tail type I collagen solution (3.4 mg/mL), 50 μL of 10 basal Eagle’s medium, and 2 μL of 7.5% sodium bicarbonate solution [10] are mixed with an automatic pipette until obtaining a homogeneous matrix of a yellowish color. The mixture is immediately used to prepare the matrices. A single drop of 30 μL is deposited on the dried coated coverslips and kept in the incubator at 37
C and 5% CO2 for at least 2 h to induce collagen gel formation (see Note 1).

3.2 Spinal Cord Organotypic Slices and DRG Explants Cultures

To increase neuronal survival, the following steps should be swiftly performed, and tissue should be kept cold. Flat ice packs can be used as a base. Postnatal rats (around p7 would be ideal) are deeply anesthetized with an overdose of pentobarbital (200 mg/kg i.p.) and decapitated. Under a light microscope, spinal cords and DRGs are harvested (see Note 2 ). Briefly, a laminectomy is performed using a special forceps to manipulate the bone of mice or postnatal rats (laminectomy forceps). The spinal cord is carefully exposed from the cervical to the sacral levels. DRGs from the desired anatomical localization (usually thoracic and/or lumbar) are harvested by picking the nerve root and cutting the sample by fine spring scissors. After harvesting the DRG, the remaining roots are also cut with the scissors to isolate the spinal cord. The spinal cord can be divided into two segments (cervical-thoracic and lumbar-sacral) to facilitate manipulation and to reduce the risk of damaging the tissue. With a fine forceps and the assistance of the scissors, the spinal cord segment is carefully released from the bone and meningeal tissue before harvesting. Samples are placed in Gey’s balanced salt solution enriched with 6 mg/mL glucose. Three washes using the same solution are performed under the cell culture hood to sterilize the samples. A portable light microscope is placed into a cell culture hood with laminar flux to clean the samples from blood, meningeal debris, and connective tissue. Samples have to be kept cold during the procedure by using an ice pack as a base. The connective tissue of the DRG and the meninges of the spinal cords should be thoroughly eliminated while maintaining the integrity of the samples (see Note 3). To prevent the damage of the spinal cord, carefully remove the meninges from the caudal to the rostral part.

3.2.1 Obtention of the Primary Tissue

434

Sara Bolı´var et al.

Spinal cords are cut in segments of about 2 cm and carefully placed on top of a round Nylon hydrophilic membrane filter with the dorsal horn facing up, then cut with a McIlwain Tissue Chopper into 350 μm thick slices. The samples are immediately placed in Petri dishes again with Gey’s balanced solution supplemented with glucose. Samples (350 μm spinal cord slices or whole DRG explants) are carefully placed on the gelled collagen droplets and a second drop of 30 μL of the same collagen solution is applied to cover them. DRG explants can be easily picked up with fine forceps whereas spinal cord slices have to be carefully collected with a small spoon. Therefore, DRG can be placed on top of collagen matrices mounted into a 24-multiwell plate, whereas it is better to use a 60 mm Petri dish with three collagen droplets to place the spinal cord slices. The embedded samples are placed again in the incubator for 45 min to allow gelation of the second droplet. Afterward, Neurobasal medium (NB) supplemented with B27, glutamine, glucose, and penicillin/streptomycin is added to each well or Petri dish. The volume added depends on the size of the recipient (1.5 mL for the 60 mm petri dish and 0.5 mL for a 24 multiwell plate). After 1 day in culture, the medium is replaced by a free-antibiotic medium (same NB-based medium, but without penicillin/streptomycin). It is recommended to culture spinal cords for 4 days to allow some spontaneous growth of motor neurites, usually less than 500 microns length (see Note 4). In contrast, DRG explants show an extensive growth already at 2 days post culture, with neurites extending about 1000 microns. During the culture, explants and spinal cord slices can be examined under an inverted light microscope. Viability of the cultures can be confirmed when migration of exogenous cells and some neurites can be observed (Fig. 1), but immunohistochemistry is needed in order to corroborate these findings and to visualize both the somas and the neurites (Fig. 2). 3.3 Modification of the Matrix

Different factors and drugs can be added directly to the matrix during its preparation, to test their effects on neuron survival and growth. Examples of agents that can be added to the matrix range from a wide variety of trophic factors like BDNF, NT3, GDNF, NGF, and FGF [11, 12], tropic factors like laminin or fibronectin [13], combination of both types of factors [14], and peptides [15] among others. Briefly, factors can be added to 50 μL of 10 basal Eagle’s medium before mixing with collagen. As a reference, addition of 10 ng/mL of classical trophic factors like GDNF or BDNF is enough to promote neurite outgrowth in the samples ([11], see also Figs. 2 and 3).

DRG and Spinal Cord Organotypic Cultures on Collagen Matrices

435

Fig. 1 Representative image of a fresh 4-day spinal cordorganotypic cultures taken from an inverted microscope (a) and magnification of the ventral horn (b). Neurite growth from the ventral horn and migration of cells from the explant into the matrix can be appreciated, indicative of a viable culture. Bar ¼ 150 μm

Fig. 2 Representative images of confocal photomicrographs of rat DRG explants and spinal cordorganotypic slices after 2 and 4 days in culture, respectively. Samples have been processed for immunocytochemistry to label myelinated fibres and their somas (by using RT97 antibody (red)). DRG explant with addition of 10 ng/mL GDNF into the matrix, where somas of myelinated sensory neurons (arrow) are labeled, as well as their neurites, that extend into the collagen matrices from all around the explant (a). Higher magnification of the explant to observe migration of cells (arrow) from the DRG to the matrix. Nuclei of cells have been labeled with DAPI (blue) (b). Spinal cord slice with addition of 10 ng/mL BDNF into the matrix. Motoneurons (arrows), with the typical triangular shape can be appreciated in the ventral horn, with neurites extending into the collagen matrix (c). Bar: 50 μm (a) and 100 μm (c)

Similarly, cells can also be included in the matrix to perform cocultures. Mesenchymal stem cells, fibroblasts, Schwann cells, or olfactory ensheathing glia [11, 16, 17] have been added successfully to the collagen matrix. The specific cells are obtained previously from primary cultures. Then, an adequate amount of cell suspension in the growing medium is carefully mixed in the collagen solution to get the final concentration. As a reference, between 10.000 and 50.000 cells can be mixed to each volume of collagen matrix used to embed the spinal cord slices or DRG explants. This number of cells is sufficient to influence the fate of the cocultures (Figs. 3 and 4, [18]).

436

Sara Bolı´var et al.

Fig. 3 Evaluation of neurite outgrowth of rat DRG explants cultured into collagen matrices seeded with 10,000 fibroblasts, Schwann cells (SC) or olfactory ensheathing glia (OEG). Column plots of neurite arborization (a) and elongation (b) when the different cells were added into the matrix. Results are expressed as mean  SEM. One-way ANOVA with post hoc Bonferroni method was used to statistically analyze neurite outgrowth, *p < 0.05 compared to fibroblasts cocultures. Representative images of DRG neurite outgrowth in cocultures with GFP transfected SCs (c) and OECs (d). In both situations, cells are well integrated in the collagen matrix and no cluster formations can be observed. RT97 in red, GFP in green, and DAPI in blue. Bar ¼ 100 μm. Both OEC and SC promote growth of primary sensory neurons. Adapted from [18] 3.4 Source of Primary Tissues

This protocol is extensively used in our lab using postnatal rats, and it has been adapted recently to postnatal mice. Since survival of postnatal mouse motoneurons can be more challenging than that of rat motoneurons, adding GDNF in control conditions is recommended in other types of spinal cord explants [19]. Addition of GDNF (10 ng/mL) into the matrix promotes considerable neurite growth in our cultures, although non-treated spinal cord slices show some growth as well (Fig. 5a and d vs. b and e). In contrast, survival of primary sensory neurons does not differ between mouse and rat neurons.

DRG and Spinal Cord Organotypic Cultures on Collagen Matrices

437

Fig. 4 Evaluation of neurite outgrowth of rat spinal cordorganotypic slices cultured into collagen matrices seeded with 10.000 fibroblasts, Schwann cells (SC) or olfactory ensheathing glia (OEG). Column plots show the mean values for motor neurite arborization (a) and elongation (b) in cocultures with the three different cell types and in control condition. Results are expressed as mean  SEM. One-way ANOVA with post hoc Bonferroni method was used to statistically analyze neurite outgrowth, *p < 0.05 vs. fibroblast conditions. Confocal pictures showing motoneuronneurite outgrowth from the spinal cord slice cocultured with 10.000 SCs (arrows indicate neurite elongation into the matrix) (c) or OECs (d). SCs appeared spread in the collagen matrix and some in relation to neurites elongating within the matrix (c). In OEC cocultures, clustering of the grafted cells (GFP+, green) can be observed surrounding the spinal cord slice. Motoneurons elongate neurites within the spinal cord slice (arrows) but not into the matrix, thus avoiding contact with the cell clusters. Therefore, OEC are creating a nonpermissive environment for growth into the matrix (d). RT97 in red, GFP in green, and DAPI in blue. Bar ¼ 100 μm (c) and 200 μm (d). Adapted from [18] 3.5 Immunohistochemistry

Spinal cord slices and DRGs embedded in the collagen matrix are fixed with PFA 4% pH 7.4 for 30 or 15 min at room temperature (RT), respectively, and then washed with TBSL 3 times (20 min each) at RT on a shaker. For the spinal cord slices, antigen retrieval is recommended, keeping samples into hot citrate buffer for at least 1 h. Then, a wash with TBSL for 5 min at RT on a shaker is performed. Next, samples

438

Sara Bolı´var et al.

Fig. 5 Representative images of a 4-day spinal cord organotypic culture of an 8-day postnatal mice in control conditions (a, d) or treated with GDNF (10 ng/mL; b, e). Extensive neurite outgrowth from the ventral horn can be observed in GDNF treated culture, although some neurites and motoneurons are also observed in the control culture (e vs. d). The samples were labeled with RT97 (a, d) or NF200 (b, c, e, f). Some motoneurons with the typical triangular shape can be observed (arrow d, e), although the soma is not as well labeled as the neurites. Example of a mice DRG explant cultured for 2 days and immunostained with NF200 (c, f). Similar to spinal cord slices, there is a fain staining of the somas, being neurites highly contrasted. Bar ¼ 150 μm

are incubated in methanol of increasing concentrations (50%, 70%, and 100% H2Od) for 20 min and washed two times with TBSL (for 20 min). Both DRGs and spinal cords are incubated with primary antibody for 48 h at 4
C in TBSL with 1.5% specific serum (see Note 5). Three washes with TBSL for 1 h are performed, and then samples are incubated with secondary antibody overnight at 4
C in TBSL with 1.5% specific serum (the same serum used for the primary incubation). Then, samples are washed with TBSL 2 times, and with TBS and TB once (for 40 min each). Samples are dehydrated in progressive ethanol concentration (70%, 96%, and 100%, for 1 min). The round coverslips are removed from the dishes with the help of a 26G needle and fine forceps and let dry. Coverslips are finally covered with the desired mounting media (for example, MOWIOL, an aqueous mounting media already prepared with DAPI (1:1000)).

DRG and Spinal Cord Organotypic Cultures on Collagen Matrices

3.6

Quantification

3.6.1 Neuronal Survival Evaluation

3.6.2 Neurite Growth Measurement

4

439

To assess the potential contribution of the different conditions tested on neuronal survival, spinal cord slices can be stained with anti-neurofilament antibody to label motoneurons and myelinated sensory neurons, and their growing neurites (see Note 6). For a fair comparison between spinal cord slices and DRG explants, both myelinated motor and sensory neurons can be labeled (Figs. 2 vs. 3, Fig. 5), although non myelinated neurons can also be stained in the DRG (see Note 7). Immunofluorescent neurons are counted at least in six samples per condition under a confocal microscope. In the spinal cord slices, only cells with morphology corresponding to motoneurons and a clear visible nucleus have to be counted in each section. For quantitative analysis of neurite growth, microphotographs are taken at 20 with a digital camera attached to an epifluorescence microscope, then automatically photomerged (when needed) and analyzed with free ImageJ software. The length of the longest neurite is measured for at least 20 samples per condition (see Note 8). To evaluate the arborization area, micrographs are transformed to 8-bit image into grayscale and quantification is performed after defining a threshold for background correction (see panels a and b from Figs. 2 and 3 for examples of quantification of both neurite length and arborization).

Conclusions The use of 3D collagen matrices allows the culture of both organotypic slices and DRG explants in a setting that is permissive for neuritogenesis and enables the measurement and comparison of positive or adverse events on neurite growth of motor and sensory neurons. Moreover, the matrix can be modified by adding trophic or tropic factors, cells, or other agents and, thus, it is a useful tool to evaluate and fairly compare neurotoxic or pro-regenerative effects in both motor and sensory neurons in a short-term culture setting.

5

Notes 1. When placing the spinal cord slices onto the collagen gel, the 24-multiwell plate can be a nuisance. Therefore, placement of the collagen matrices in a small Petri dish would facilitate the task. A 60 mm Petri-dish can contain 3 coverslips and, therefore, 3 matrices. In contrast, DRG explants are easy to manipulate with fine forceps and their placement in a 24-multiwell plate is practical.

440

Sara Bolı´var et al.

2. Some authors prefer to obtain slices from thoracic segments of the spinal cord and DRG explants from the same thoracic level of the cord since these are more uniform. In our hands, slices obtained from lumbar and sacral segments of the spinal cord can also be cultured, with the sacral ones showing higher neuronal survival rate, probably due to smaller size and more favorable oxygenation conditions. It is important to consider the heterogeneity of the samples when comparing different treatments. Similarly, lumbar DRGs are bigger, and therefore the observed growth can be different than the one observed in the more spherical thoracic ones. 3. For the spinal cords, it is important to clean the tissue from the meningeal layers, both the external and the internal one. If some meningeal debris are kept, the slices can get attached to the Nylon membrane filter during the cutting process with the chopper. 4. This type of culture is designed for short-term survival of organotypic slices, less than a week, since oxygenation of the tissue is limited. For longer survival, the use of membrane culture inserts (e.g., Millicells) is needed. Spinal cord slices are placed on these permeable membranes, to avoid direct contact with the culture media, allowing long-term survival of the tissue. However, motor neurites do not have a substrate to grow, and evaluation of neurite growth in this setting is trickier than into a matrix. Moreover, there is no equivalent culture for DRG explants and, thus, there is not a fair comparison between motor and sensory neurite growth when using these membrane inserts to culture spinal cord slices. 5. Incubation times are quite long due to the thickness of the spinal cord slices. DRG explants can be incubated half of the time. 6. Primary antibodies that are good markers for motoneurons and myelinated sensory neurons in rat samples are mouse RT97 (1:200, Developmental Studies Hybridoma Bank), mouse SMI32 (1:2500, Sternberger Monoclonals Inc.), and chicken anti-NF200 (1:1000, Millipore). It is important to note that phosphorylated neurofilament (RT97) labels mainly neurites, and non-phosphorylated one (SMI32) the somas of intact motoneurons. Since the phosphorylate form can also be detected in the soma after injury [11], this form is our preferred choice to label these cultures. Unfortunately, in mice cultures, immunolabeling against R797 in somas is weaker than the one observed in rats (compare Fig. 5 vs. Fig. 2). 7. The pan-neuronal marker PGP is a good antibody to label all the neurons of the DRG, but then neurite growth and survival would also include unmyelinated populations of neurons, whereas motoneurons are exclusively myelinated.

DRG and Spinal Cord Organotypic Cultures on Collagen Matrices

441

8. Control spinal cord slices are viable but do not always have neurite growth after 4 days. Therefore, it is important to take into account that basal neurite growth can be close to zero in control conditions. When evaluating adverse effects on neurite growth, an optimized control has to be considered (for example, addition of cells or classical trophic factors into the culture). References 1. Argyriou AA, Briani C, Cavaletti G et al (2013) Advanced age and liability to oxaliplatininduced peripheral neuropathy: post hoc analysis of a prospective study. Eur J Neurol 20 (5):788–794 2. Lehmann HC, Staff NP, Hoke A (2019) Modeling chemotherapy induced peripheral neuropathy (CIPN) in vitro: prospects and limitations. Exp Neurol 326:113140. https:// doi.org/10.1016/j.expneurol 3. Tucker BA, Mearow KM (2008) Peripheral sensory axon growth: from receptor binding to cellular signaling. Can J Neurol Sci 35 (5):551–566 4. Rothstein JD, Jin L, Dykes-Hoberg M, Kuncl RW (1993) Chronic inhibition of glutamate uptake produces a model of slow neurotoxicity. Proc Natl Acad Sci U S A 90(14):6591–6595 5. Kosuge Y, Sekikawa-Nishida K, Negi H et al (2009) Characterization of chronic glutamatemediated motor neuron toxicity in organotypic spinal cord culture prepared from ALS model mice. Neurosci Lett 454(2):165–169 6. Montoya GJ, Sutachan JJ, Chan WS et al (2009) Muscle-conditioned media and cAMP promote survival and neurite outgrowth of adult spinal cord motor neurons. Exp Neurol 220:303–315 7. Pandamooz S, Salehi MS, Zibaii MI, Safari A, Nabiuni M, Ahmadiani A, Dargahi L (2019) Modeling traumatic injury in organotypic spinal cord slice culture obtained from adult rat. Tissue Cell 56:90–97 8. Guzma´n-Lenis MS, Navarro X, Casas C (2009) Drug screening of neuroprotective agents on an organotypic-based model of spinal cord excitotoxic damage. Restor Neurol Neurosci 27(4):335–349 9. Herrando-Grabulosa M, Mulet R, Pujol A et al (2016) Novel neuroprotective multicomponent therapy for amyotrophic lateral sclerosis designed by networked systems. PLoS One 11 (1):e0147626 10. Tucker A, Lumsden A, Guthrie S (1996) Cranial motor axons respond differently to the floor plate and sensory ganglia in collagen gel co-cultures. Eur J Neurosci 8(5):906–916

11. Allodi I, Guzma´n-Lenis MS, Herna`ndez J et al (2011) In vitro comparison of motor and sensory neuron outgrowth in a 3D collagen matrix. J Neurosci Methods 198:53–61 12. Allodi I, Casals-Dı´az L, Santos-Nogueira E et al (2013) FGF-2 low molecular weight selectively promotes neuritogenesis of motor neurons in vitro. Mol Neurobiol 47:770–781 13. Gonzalez-Perez F, Ale´ A, Santos D et al (2016) Substratum preferences of motor and sensory neurons in postnatal and adult rats. Eur J Neurosci 43:431–442 14. Santos D, Gonza´lez-Pe´rez F, Giudetti G et al (2017) Preferential enhancement of sensory and motor axon regeneration by combining extracellular matrix components with neurotrophic factors. Int J Mol Sci 18(1):65. https://doi.org/10.3390/ijms18010065 15. Auer M, Allodi I, Barham M et al (2013) C3 exoenzyme lacks effects on peripheral axon regeneration in vivo. J Peripher Nerv Syst 18:30–36. https://doi.org/10.1111/jns5. 12004 16. Allodi I, Mecollari V, Gonza´lez-Pe´rez F et al (2014) Schwann cells transduced with a lentiviral vector encoding Fgf-2 promote motor neuron regeneration following sciatic nerve injury. Glia 62:1736–1746 17. Torres-Espı´n A, Corona-Quintanilla DL, Fore´s J et al (2013) Neuroprotection and axonal regeneration after lumbar ventral root avulsion by re-implantation and mesenchymal stem cells transplant combined therapy. Neurotherapeutics 10(2):354–368. https://doi.org/10. 1007/s13311-013-0178-5 18. Allodi I (2012) Changing the intrinsic growth capacity of motor and sensory neurons to promote axonal growth after injury. Thesis dissertation. https://www.tdx.cat/handle/10803/ 96355 19. Rakowicz WP, Staples CS, Milbrandt J et al (2002) Glial cell line-derived neurotrophic factor promotes the survival of early postnatal spinal motor neurons in the lateral and medial motor columns in slice culture. J Neurosci 22 (10):3953–3962

Chapter 20 Integrative In Vitro/Ex Vivo Assessment of Dopaminergic Neurotoxicity in Rodents Using Striatal Synaptosomes and Membrane Preparations Rau´l Lo´pez-Arnau and David Pubill Abstract The striatum is a brain area with a high density of dopaminergic terminals and so it is considerably affected when dopaminergic neurotoxicity occurs. Several methods have been developed to evidence such neurotoxicity, some of which can be performed in vitro, using membranes or synaptosomes obtained from fresh tissues. Concretely, here we detail our methodological experience assessing reactive oxygen species in striatal synaptosomes, as well as measuring brain terminal damage after a neurotoxic treatment leading to decreased [3H]WIN 35428 binding and tyrosine hydroxylase expression measured by Western blot. Also, we have assessed impairment of dopamine transporter after dopaminergic neurotoxicity measuring uptake of [3H]dopamine in striatal synaptosomes. These techniques are complementary and can be useful to assess the dopaminergic neurotoxic potential of drugs such as amphetamine derivatives, among others. Key words Dopamine uptake, Dopaminergic neurotoxicity, Reactive oxygen species, Striatum, Synaptosomes, Tyrosine hydroxylase, [3H]WIN 35428 binding

1

Introduction It is well known that neurotoxicity occurs when a physical, chemical, or biological agent produces, directly or indirectly, any long- or short-term adverse effect in the central or peripheral nervous system structure or function [1]. In this context, the dopaminergic system plays a key role in the neurotoxic effects observed in the central nervous system (CNS) following abuse of certain drugs (e.g., amphetamines), leading to neurochemical impairments which, to a certain extent, can resemble those of diseases such as Parkinson’s (PD). Synaptosomes are synaptic terminals isolated from a certain brain area which keep the morphological features and most of the chemical properties of the original nerve terminal [2]. They contain numerous small clear synaptic vesicles, sometimes larger dense-core

Jordi Llorens and Marta Barenys (eds.), Experimental Neurotoxicology Methods, Neuromethods, vol. 172, https://doi.org/10.1007/978-1-0716-1637-6_20, © Springer Science+Business Media, LLC, part of Springer Nature 2021

443

444

Rau´l Lo´pez-Arnau and David Pubill

vesicles, and frequently one or more mitochondria. For these reasons, synaptosomes are used to study synaptic transmission in vitro because they contain the molecular machinery necessary for the neuronal uptake, storage, and release of neurotransmitters. They maintain a normal membrane potential, contain presynaptic receptors, translocate metabolites and ions, and when depolarized, release multiple neurotransmitters (including catecholamines) in a calcium-dependent manner. The striatum is a brain area with predominant dopaminergic innervation, which allows studying effects related with dopamine (DA) (i.e., uptake, storage, and release) but striatal synaptosomes can also be used to assess oxidative stress measuring reactive oxygen species (ROS) induced by substances at the nerve terminal, which is indicative of possible neurotoxicity. The works carried out by Myhre and Fonnum [3] on the neurotoxic effects of hydrocarbons and our studies on the oxidative effects of amphetamine derivatives [4–8] are good examples. Many amphetamine derivatives (e.g., methamphetamine, 3,4-methylenedioxy-methamphetamine (MDMA)) are taken up into the nerve terminals and displace monoamines (DA, serotonin) from their storage vesicles, inducing their release through reverse transport. However, this effect is followed by a blockade of the transporter and free monoamines can undergo cytosolic metabolism producing several types of ROS (i.e., hydrogen peroxide, hydroxyl radical, peroxynitrite, quinones) which react with cell structures and produce neurotoxicity [9–11]. Thus, measuring increased ROS production in nerve synaptosomes is indicative of neurotoxicity. As DA is the predominant neurotransmitter in the striatum, ROS production in this area presumes dopaminergic neurotoxicity. Moreover, the fact of being an in vitro preparation also allows studying the involved mechanisms in an isolated and composition-controlled medium by the addition of specific drugs acting at receptors, inhibiting enzymes, etc. Some techniques involve functional imaging of dopamine transporter (DAT) in human striatum by positron emission tomography or single photon emission computerized tomography using different radiolabeled DAT ligands (e.g., cocaine or WIN 35428). In fact, DAT imaging is used for detection, for example, of early stage PD [12, 13]. Moreover, neuroimaging studies, which reveal a decrease in DAT levels, have demonstrated the neurotoxic effects in humans induced by some amphetamine derivatives, leading to memory and motor impairments [14]. Changes in DAT expression have been also involved in other human brain disorders, such as dementia with Lewy bodies, Wilson’s disease, Machado–Joseph disease, Lesch–Nyhan disease, Gilles de la Tourette’s syndrome, and schizophrenia [12, 15, 16]. In addition, and along with some neurodegenerative changes, a reduction in DA uptake and a decrease in tyrosine hydroxylase (TH) levels, the rate-limiting

In vitro/ex Vivo Assessment of Dopaminergic Neurotoxicity in Rodents

445

enzyme of the DA biosynthesis, have been also observed in studies with animals [17, 18] (see also review [19]) and are broadly used as markers of dopaminergic impairment. Here we show the protocols we used in our experiments, which were focused on studying the dopaminergic neurotoxicity of amphetamine derivatives such as methamphetamine, MDMA, and mephedrone, among others.

2

Animals, Materials, Buffers, and Reagents Protocols involving laboratory animals must be compliant with local and national animal welfare regulations and be approved by the corresponding Ethics Committees. Although the techniques explained in this chapter can be adapted to other species, the methods we report have been applied to rats and mice. Rats and mice from the most popular strains can be used. We mostly used Sprague Dawley and Dark Agouty rats, as well as Swiss-CD-1 mice. The age will depend on the study design. For experiments involving in vivo drug treatment and ex vivo determinations, no less than six animals per group should be used. For experiments on preparations from drug-naı¨ve animals, please see Subheading 3.1.1. All the reagents should be of analytical grade and some can be obtained from the most habitual commercial sources. – Tris-sucrose buffer: 5 mM Tris–HCl and 320 mM sucrose. It can be prepared by diluting an aliquot of a stock 50 mM Tris– HCl (pH 7.4, store at 4
C) in bidistilled water and adding the sucrose. Dissolve by stirring and adjust pH to 7.4. It is advisable to freshly prepare it, as the presence of sucrose may favor growth of microorganisms. – Tris-sucrose 1.6 M buffer: prepare as described above but using the corresponding amount of sucrose. – PBS-sucrose buffer: it consists of sodium phosphate-buffered (100 mM) solution with sucrose 320 mM at pH 7.9. – HBSS-glucose buffer: it consists of the standard Hank’s Balanced Saline Solution (HBSS) plus 20 mM HEPES sodium and 5.5 mM glucose. HBSS composition is (in mM): 140 NaCl, 5.37 KCl, 1.26 CaCl2, 0.44 KH2PO4, 0.49 MgCl2, 0.41 MgSO4, 4.17 NaHCO3, 0.34 Na2HPO4. This buffer can be purchased from some commercial sources (e.g., Biological Industries, Inc.) and HEPES and glucose should be freshly added to the desired volume. Alternatively, HBSS plus HEPES can be prepared in the laboratory as a concentrated stock (e.g., 2) and the desired volume can be diluted, and the glucose added the day of the experiment. HBSS should be stored at 4
C.

446

Rau´l Lo´pez-Arnau and David Pubill

– 20 ,70 -Dichlorofluorescin diacetate (DCFH-DA). It is the reagent that will produce fluorescence after being taken up by the synaptosomes and react with ROS. Molecular Probes is a habitual source to purchase it. It is a powder which can be previously dissolved and aliquoted as convenient to perform the experiments. Typically, dissolve 10 mg in 2 mL of anhydrous analytical-grade DMSO. This gives a 10 mM stock. Minimize light exposure during this process. Store in aliquots of 100 μL or as convenient in eppendorfs at 20
C, protected from light. Adding 5 μL of this solution to 1 mL of synaptosome suspension will provide the desired final concentration of 50 μM. – Radioligands: [3H]DA (DAT substrate for uptake experiments) and [3H]WIN 35428 (DAT ligand, for binding experiments) can be purchased from Perkin Elmer (Boston, MA, USA). [3H] DA stock is stored frozen and at 4
C once thaw. [3H]WIN 35428 ethanolic solution must be stored at 20
C. Both compounds must be diluted previously as described in the corresponding Subheading 3. – Antibodies and reagents for Western blot: mouse monoclonal antibody against TH (Cat. N
: 612300 Transduction Labs, Lexington, KY, USA), peroxidase-conjugated anti-mouse IgG antibody (GE Healthcare, Buckinghamshire, UK), beta-actin (mouse monoclonal antibody, Sigma-Aldrich), and chemiluminescence-based detection kit (Immobilon Western, Millipore, USA) are used for TH detection. Equivalent reagents from other brands should also work. – Western blotting itself would merit a complete chapter as many variations can be performed. For this reason, please refer to [20– 22] for general methodology and apparatus required (e.g., tanks, cassettes, power supply) and we will explain the particularities for TH determination in the corresponding section. All the reagents used for preparing the buffers should be of electrophoresis-reagent quality. – Other materials: scintillation liquid (e.g., Ultima Gold, Perkin Elmer), Bio-Rad Protein Reagent for total protein determination (Bio-Rad Labs. Inc., Hercules, CA), GF/B glass fiber filters (Whatman, Maidstone, UK), polyvinylidene fluoride (PVDF) sheets (Immobilon-P, Millipore, Billerica, MA, USA), vacuum manifold (or Cell Harvester), scintillation counter (beta), borosilicate glass homogenizing tube, standard glass tubes ( 6 mL) (see Note 1), and scintillation plastic vials.

In vitro/ex Vivo Assessment of Dopaminergic Neurotoxicity in Rodents

3

447

Methods

3.1 Measurement of ROS Production in Striatal Synaptosomes 3.1.1 Obtention of Striatal Synaptosomes for ROS Measurement

Although the P2 fraction, a crude synaptosome preparation whose obtention is described in Subheading 3.3, can be used for many purposes (e.g., uptake experiments), measuring ROS production requires a more purified preparation devoid of free mitochondria which could generate ROS outside the synaptosomes and provide inconsistent effects. Several methods are available to purify synaptosomes, including Percoll gradient centrifugation which is delicate to apply. However, the protocol we use is based on that described by Myhre and Fonnum [3] and allows enough purity of the preparation to perform such determinations (Fig. 1). Basically, the modification of the protocol consists in increasing the sucrose concentration when pelleting the synaptosomes, so that the free mitochondria layer separates from the synaptosomes. The protocol to be used is as follows: The whole experiment must be carried out within the same day, with fresh samples. Freezing of synaptosomes results in loss of some of their functional properties. 1. Animals: for a typical experiment, two rats (Sprague Dawley weighing 250–325 g, but other strains should work) or 7–8 mice (30–35 g, we used Swiss-CD-1 but others should work) (see Note 2) provide enough material for a typical experiment. Rats are sacrificed by decapitation under isoflurane anesthesia, while mice can be sacrificed by rapid cervical dislocation followed by decapitation.

Fig. 1 Representative electron microscopy picture of the synaptosomal fraction obtained by the method described (X 65,000). Reproduced from Escubedo et al., 2009, with permission from Elsevier

448

Rau´l Lo´pez-Arnau and David Pubill

2. Dissection of the striatum: immediately after the sacrifice, the skull must be opened, the brain removed and put on a refrigerated surface (hint: a sheet of modeling wax on an icepack provides the appropriate temperature and avoids tissue sticking to the surface). Using thin forceps, separate the cortex following the cerebral longitudinal fissure and cut the white fibers just below (corpus callosum). Separate the two hemispheres and the striata can be seen in the middle of the anterior moiety as two spheroidal corpora. Excise carefully using the forceps and keep in a glass Petri dish or a beaker on ice while processing the rest of animals and until the next step. 3. Weigh the striata and homogenize them in 20 volumes of Trissucrose buffer (weight in g x 20 mL/g) using a motor-driven Teflon/glass homogenizer (6 strokes at approximately 1600–1800 rpm). Keep the tube on ice during the homogenization to avoid heating. 4. Filter the homogenate through two layers of surgical gauze in order to retain eventual unbroken tissue. Transfer the homogenate in two equal volumes to centrifuge tubes. 5. Centrifuge at 1000  g at 4
C in a refrigerated centrifuge for 10 min. 6. Recover the supernatant with a Pasteur pipette and transfer it to a tube or small beaker (keep always on ice). Then, measure the volume (V) of the supernatant using a micropipette while transferring it to a clean centrifuge tube. Add V  0.6 mL of cold Tris-sucrose 1.6 M buffer, cap the tube, and mix well by repetitive inversion, so that the final sucrose concentration will be 0.8 M. Split it into two clean centrifuge tubes. 7. Centrifuge at 13,000  g for 30 min at 4
C. Then, discard the myelin-rich supernatant. The pellet will appear as a bottom brown-colored layer consisting in mitochondria, covered by a white layer of synaptosomes. 8. Carefully add 1 mL of cold Tris-sucrose buffer sliding along the tube walls and resuspend the white layer by gently shaking or tapping the tube. Avoid resuspending the brown layer. Recover the resuspended synaptosomes with a Pasteur pipette and transfer them to a tube (e.g., a conic 50 mL Falcon-type tube). 9. Dilute the synaptosomes with HBSS-glucose buffer up to a volume around 18 mL. This should give a final protein concentration around 0.1 mg/mL. Keep on ice until starting the incubation steps. 10. This suspension can be distributed in 1 mL aliquots into centrifuge tubes. This is advisable as further centrifugations will be performed after incubation. We prefer using regular centrifuge tubes than eppendorfs, as they permit better shaking of the suspension in the incubation bath.

In vitro/ex Vivo Assessment of Dopaminergic Neurotoxicity in Rodents 3.1.2 Incubation with Drugs and Fluorochrome

449

ROS production is measured by using the fluorochrome 20 ,70 -dichlorofluorescin diacetate (DCFH-DA) which passively diffuses through membranes and, after being deacetylated by esterases, is accumulated inside the synaptosomes in the form of 20 ,70 -dichlorofluorescine, which is not fluorescent. This compound reacts quantitatively with oxygen species to produce the fluorescent dye 20 ,70 -dichlorofluorescein (DCF), whose intensity can be measured to provide an index of oxidative stress. 1. Add 5 μL of DCFH-DA solution to each tube to reach a final concentration of 50 μM. At the same time, modulatory drugs (e.g., antagonists, enzyme inhibitors) are also added to the suspension if needed at a volume of 5–10 μL (see Note 3). Remember to always perform a control tube of synaptosomes which only will receive the fluorochrome, as it will provide the basal ROS production (see Note 4). Gently vortex each tube for 2 s. Avoid direct exposure to light which would degrade the fluorochrome, so work under low-light conditions and incubate in the dark or alternatively, with the tube rack covered by aluminum foil. 2. Incubate in a thermostatic shaking bath at 37
C for 15 min at moderate speed (e.g., 70 shakes/min). During this time, synaptosomes will take up the fluorochrome and the modulatory drugs will equilibrate with its target. 3. Add the oxidative stimulus to test. Use small amounts (e.g., 5–10 μL) to not significantly increase the final volume. A positive control can also be performed in a tube adding 100 μM H2O2. Moreover, a nonspecific antioxidant effect of any of the modulatory compounds used can be assessed in tubes containing the compound at the concentration used and 100 μM H2O2. 4. Gently vortex and incubate for the final desired time. A preliminary time-course experiment is advisable, with times ranging between 30 min and 2 h. More extended incubation times are not recommended as they may compromise the viability of the synaptosomes. 5. Once the incubation is finished, centrifuge at 13,000  g for 20 min at 4
C. Discard the supernatant and resuspend the pellet with 1 mL Tris-sucrose buffer by gently vortexing. This step will wash out residual drugs. Repeat the centrifugation and resuspend the final pellets with 0.2 mL HBSS-glucose buffer. Keep the samples on ice and protected from direct light until fluorescence measurements are carried out.

3.1.3 ROS Measurement

ROS-induced fluorescence can be measured by means of a flow cytometer equipped with an argon laser at an excitation wavelength of 488 nm and detecting emission at 525 nm (see Note 5). The advantage of using this equipment is that synaptosomes can be

450

Rau´l Lo´pez-Arnau and David Pubill

Fig. 2 Representative flow cytometry histograms (a) and dot plots (b, c) showing the change in DCF fluorescence of rat striatal synaptosomes after 2-h incubation at 37
C alone (Control, CTRL) or with 2 mM methamphetamine (METH). Reproduced from Escubedo et al., 2009, with permission from Elsevier

separated in different populations according to their shape/size and fluorescence, so that different analysis possibilities are available. To perform the measurements, dilutions of the suspension must be made in order to achieve a flow rate of 500–900 synaptosomes/ s and each sample should be measured for 1 min. This flow rate is usually achieved by diluting 10 μL of synaptosome suspension in 1 mL of HBSS/glucose buffer. It is advisable to measure triplicates of each sample. Processing of the cytometry data using dedicated software allows generating histograms and dot plots showing the displacement of the fluorescence of the synaptosomal population by effect of the treatment (Fig. 2). Also, the mean fluorescence is given in order to perform comparisons among treatment groups. Obtain the basal fluorescence by calculating the mean of the control samples. This value will be assigned the reference value of 1 or 100. Normalize the rest of the data with respect to this value for each experiment. Alternatively, total fluorescence can be assessed transferring the samples to a multiwell plate with black walls and transparent

In vitro/ex Vivo Assessment of Dopaminergic Neurotoxicity in Rodents

451

bottom and use a fluorimeter to measure fluorescence using an excitation wavelength of 488 nm and an emission wavelength of 525 nm [3, 23, 24]. 3.2 Radioligand Binding and Western Blot Assays 3.2.1 Preparation of Striatal and Cortical Crude Membranes and Cytosolic Fraction from Rodent Brains

Immediately after sacrifice by cervical dislocation and decapitation (mice) or by decapitation under isoflurane anesthesia (rats), the brains are rapidly removed from the skull. Striatum and/or frontal cortex are quickly dissected out on an ice-cold surface as described above, but then they must be frozen on dry ice, and can be stored at 80
C until use. 1. The day of the experiment, tissue samples are weighed, thawed, and homogenized with a sonicator homogenizer at cycle mode of 0.95 (for the sonication of heat-sensitive tissue) and amplitude of 95–100 for 20 s (mice) or with a Polytron homogenizer, at around 12,000 rpm, for 15 s (rats) at 4
C in 10 (mice) or 20 (rats) volumes of Tris-sucrose buffer containing protease inhibitors (aprotinin 4.5 μg/μL, 0.1 mM phenylmethylsulfonyl fluoride, and 1 mM sodium orthovanadate) at pH 7.4. 2. The homogenates are centrifuged at 1000  g for 15 min at 4
C. Aliquots (100 μL) of the resulting supernatants can be stored at 80
C until use for Western blot experiments. This strategy allows using the tissues from the same animals for both WB and binding assays. If membranes for binding are not going to be used, total tissue lysates in, for example, RIPA buffer [20–22] can also be used for WB experiments. 3. The content of the tube is mixed again by vortexing and centrifuged at 15,000  g for 30 min at 4
C. 4. The resulting pellets are resuspended (in the same volume used in step 1.) in cold Tris-sucrose buffer with protease inhibitors and incubated at 37
C for 5–10 min to remove endogenous neurotransmitters that could interfere the determinations. 5. Samples are centrifuged at 15,000  g for 30 min at 4
C and resuspended again in the same volume as above. Repeat this centrifugation. 6. Discard the supernatant and resuspend the final pellets in 0.4 or 0.75 mL for mice and 1 or 1.5 mL for rats (see Note 6) of PBS-sucrose buffer for striatal or cortical membrane preparation, respectively. 7. Reserve an aliquot (20–30 μL) of each sample to determine protein content and store the rest at 80
C until use in radioligand binding experiments.

3.2.2 [3H]WIN 35428 Binding Assay in Crude Membrane Preparation

The density of DAT in striatal or frontal cortex membranes from mouse or rat can be measured by [3H]WIN 35428 binding assays, as a marker of dopaminergic nerve terminals (see Fig. 3).

452

Rau´l Lo´pez-Arnau and David Pubill

Fig. 3 Effect of a mephedrone (Meph) exposure (3 doses of 25 mg/kg, s.c. at 2 h interval for 2 consecutive days) on DAT density in mouse striatum (a) and frontal cortex (b). Results are expressed as mean  SEM from 8 to 10 animals in percentage of control (saline). *p < 0.05, **p < 0.01, and ***p < 0.001 vs. saline. Authors found a relationship between the decrease in the [3H]WIN 35428 specific binding and the decrease in TH expression in the frontal cortex (saline: 100.00  2.46%; mephedrone: 60.04  9.34% p < 0.01, 3 days after exposure and 55.30  9.11% p < 0.001, 7 days after exposure). Modified from Martı´nez-Clemente et al., 2014 [27]

1. Crude membrane preparation (see Subheading 3.2) must be previously diluted in PBS-sucrose buffer at 4
C to a concentration of 1 μg/μL. This buffer has been demonstrated to enhance the specific binding of the radioligand [25]. 2. The experiment can be performed in borosilicate glass tubes as follows: (a) Total binding tubes; 50 (striatum) or 100 μL (frontal cortex) of crude membranes preparation previously diluted (50 μg or 100 μg, respectively) and 150 (striatum) or 300 μL (frontal cortex) of PBS-sucrose buffer. (b) Nonspecific binding tubes; contain the same as total binding tubes but also bupropion (30 μM final concentration) is added so that all the specific binding sites are occupied by this cold ligand and radioligand only will be able to bind nonspecific sites under these conditions. Prepare at least duplicates for each condition (sample/total and nonspecific binding tubes). 3. Start the incubation with the addition of 50 (striatum) or 100 μL ( frontal cortex) of [3H]WIN 35428 diluted in PBS-sucrose buffer so that a final radioligand concentration of 5 nM is reached in the binding tubes. 4. Carry out the incubation for 2 h at 4
C by putting the tubes in a rack on ice. Thereafter, rapidly filter the content of the tubes under vacuum (Vacuum manifold or cell harvester) through GF/B glass fiber filters previously soaked (30 min minimum) in 0.5% polyethyleneimine solution.

In vitro/ex Vivo Assessment of Dopaminergic Neurotoxicity in Rodents

453

5. Rapidly wash tubes and filters three times with 4 mL of ice-cold Tris–HCl 50 mM buffer. Transfer the filters to plastic scintillation vials and add 5–7 mL of scintillation liquid. 6. At least 4 h after adding scintillation liquid, measure radioactivity in the filters by means of a beta scintillation counter. Calculate the means of each set of replicates and the specific binding by subtracting nonspecific binding from total binding values. Then perform the appropriate statistical analysis between groups. 3.2.3 TH Expression in Cytosolic Fraction

In order to determine the expression of TH in rodents, a general Western blotting and immunodetection protocol can be performed as described elsewhere [26]. 1. For each sample (cytosolic fraction obtained from Subheading 3.2.), mix the appropriate amount with sample buffer, boil for 5 min and load 15–20 μL containing 15 μg of protein onto a gel consisting in an upper 5% polyacrylamide stacking part followed by the 10% polyacrylamide separating gel. 1 mmthick gel is enough. 2. Run the electrophoresis at a constant voltage of 100 V until the protein front crosses the stacking gel and continue electrophoresis at constant intensity of 35 mA until elution of the protein front (blue line). 3. Transfer the proteins to PVDF sheets using a wet transfer tank for 2 h at 0.2 A. 4. Block the PVDF membranes by incubating overnight with 5% defatted milk in TBS-Tween buffer. 5. Then, incubate the membranes for 2 h at room temperature with a primary mouse monoclonal antibody against TH diluted in 5% defatted milk in TBS-Tween buffer (dilution 1:5000). Incubation can be carried out in a small tray or in a capped cylindric tube so that a moderate incubation volume can be used (e.g., 5 mL). Incubation and washing processes must be performed under continuous shaking (e.g., on an orbital shaker or on a tube roller). 6. After washing the membranes with TBS-Tween buffer (3  10 min), incubate with a peroxidase-conjugated antimouse IgG antibody diluted in TBS-Tween buffer (dilution 1:2500) for 45 min at room temperature. 7. Wash with TBS-Tween (3  10 min), add 1–2 mL of reagent from a chemoluminescence-based detection kit following the manufacturer’s protocol, and visualize and scan immunoreactive proteins using a luminescence-based gel documentation system (e.g., BioRad ChemiDoc). Scanned blots can then be

454

Rau´l Lo´pez-Arnau and David Pubill

Fig. 4 Representative Western blot for TH in striatum from mice treated with saline or MDMA and sacrificed 7 days after the treatment. The corresponding β-actin blots are shown below and were used to normalize the values and correct deviations in total protein loading. MDMA (25 mg/kg, s.c.) was given every 3 h, for a total of three doses. The mice were maintained in an environmental temperature of 26  2
C and were kept under these conditions until 1 h after the last dose. Adapted from Chipana et al., 2006 [5], with permission from Elsevier

analyzed using dedicated software (e.g., BioRad Image Lab) and dot densities expressed as a percentage of those taken from the controls (see Fig. 4). 8. Beta-actin is a cytosolic protein whose expression is not affected by drug treatments, so its immunodetection in the same membrane serves as a control of loading uniformity for each lane and is used to correct differences in TH expression due to protein content. The incubation with the anti-beta-actin antibody (mouse monoclonal, dilution 1:2500) is carried out for 1 h at room temperature. After washing the membranes with TBS-Tween buffer (3  10 min), incubate for 30 min at room temperature with a peroxidase-conjugated anti-mouse IgG antibody (dilution 1:2500), wash again (4  5 min) and visualize and quantify the chemiluminescence as described above. 9. Normalize each TH value dividing by its corresponding betaactin. Calculate the means for each treatment group. The values for control group can be set to 100% or 1 and the rest normalized accordingly. Then perform the appropriate statistical analysis. 3.3 Obtention of Striatal Synaptosomes for Uptake Experiments

As mentioned before, crude P2 fraction from synaptosomal preparation can be used for monoamine uptake experiments. The procedure is similar to that described before in Subheading 3.2, with some modifications: 1. Follow the steps 1–5 from Subheading 3.2. 2. After centrifugation, recover the whole supernatant and centrifuge at 13,000  g for 30 min at 4
C.

In vitro/ex Vivo Assessment of Dopaminergic Neurotoxicity in Rodents

455

3. Thereafter, discard the supernatant and dilute the pellet in 3–18 mL (see Note 6) HBSS-glucose buffer containing pargyline (20 μM) and ascorbic acid (1 mM). Keep on ice until starting the uptake experiment. 3.3.1 Plasmalemmal [3H] DA Uptake Experiments

In order to study the plasmalemmal functionality/density of DAT, [3H]DA uptake assays can be carried out as follows: 1. Synaptosomes from rat striatum are prepared as described above. 2. A thermostatic water bath with shaking is needed. Fill it to a level ensuring that the bottom of the tubes is immersed but they do not float and set it to 37
C with gentle shaking. Use always an appropriate tube rack and keep all the cautions needed when handling radioactive materials. 3. The experiment should also be performed in borosilicate glass tubes as follows: (a) Total uptake tubes: 125 μL of HBSS-glucose buffer (containing 20 μM pargyline and 1 mM ascorbic acid) and 100 μL of the corresponding synaptosomal suspension. (b) Nonspecific uptake tubes: composed by the same as above but the buffer contains cocaine for a final concentration of 300 μM. As these will be incubated at 4
C, specific uptake should already be blocked, so adding cocaine would be optional. We prefer adding it to ensure the complete blockade of specific uptake, but both options can be found in the literature. 4. Incubation starts with the addition of 25 μL of [3H]DA diluted in HBSS-glucose buffer (containing pargyline and ascorbic acid) for a final concentration of 5 nM to each tube. The three components, buffer, synaptosomal suspension, and [3H]DA stock solution, must be previously warmed separately for 5 min at 37
C before mixing. 5. The addition of [3H]DA starts the reaction and then the incubation in total uptake tubes is carried out for a further 5 min at 37
C. 6. For nonspecific uptake, incubate the corresponding tubes at 4
C, also for 5 min. 7. Then, terminate the uptake reaction by rapid vacuum filtration through Whatman GF/B glass fiber filters (as described in Subheading 3.2.1.) presoaked in 0.5% polyethyleneimine. 8. Rapidly wash tubes and filters three times with 4 mL ice-cold 50 mM Tris–HCl buffer. 9. Transfer the filters into plastic scintillation vials and add 5–7 mL of scintillation liquid.

456

Rau´l Lo´pez-Arnau and David Pubill

10. After at least 4 h, measure radioactivity in the filters using a liquid scintillation counter. Each condition (total and nonspecific uptake tubes) must be performed at least per duplicate. As for binding assays, specific uptake is defined as the difference between the radioactivity measured in the absence (total uptake) and in the presence (nonspecific uptake) of an excess of unlabeled DAT blocker (cocaine). Then, calculate the means for each group and perform the appropriate statistical analysis.

4

Notes and Troubleshooting 1. Other tubes or recipients may work (i.e., eppendorfs or plastic tubes) but some radioligands can adhere to plastic, so make sure that this is not your case if plastic is going to be used. 2. Other strains and from different age or weight can be used; however, different total amount of protein (membranes or synaptosomes) must be expected. 3. Some drugs need to be dissolved in DMSO. In this case, prepare stocks so that, after dilution, final concentration of the solvent does not exceed 0.5%. 4. Remember to always perform a control tube to assess basal ROS production. 5. Synaptosomes are not cells, so ask advice to the cytometer technician, who will set up the apparatus properly in order to detect the synaptosomes and will define the appropriate populations to obtain reliable results. 6. Different final volumes can be used in order to increase/ decrease protein concentration, depending on how many experiments, conditions, or experimental tubes are required.

References 1. Erinoff L (1995) General considerations in assessing neurotoxicity using neuroanatomical methods. Neurochem Int 26:111–114. https://doi.org/10.1016/0197-0186(94) 00105-4 2. Evans GJO (2015) Subcellular fractionation of the brain: preparation of synaptosomes and synaptic vesicles. Cold Spring Harb Protoc 2015:462–466. https://doi.org/10.1101/ pdb.prot083469 3. Myhre O, Fonnum F (2001) The effect of aliphatic, naphthenic, and aromatic hydrocarbons on production of reactive oxygen species and reactive nitrogen species in rat brain

synaptosome fraction: the involvement of calcium, nitric oxide synthase, mitochondria, and phospholipase A. Biochem Pharmacol 62:119–128. https://doi.org/10.1016/ s0006-2952(01)00652-9 4. Escubedo E, Chipana C, Camarasa J (2005) Methyllycaconitine prevents methamphetamine-induced effects in mouse striatum : involvement of α7 nicotinic receptors. J Pharmacol Exp Therap 315:658–667. https://doi.org/10.1124/jpet.105.089748 5. Chipana C, Camarasa J, Pubill D, Escubedo E (2006) Protection against MDMA-induced dopaminergic neurotoxicity in mice by

In vitro/ex Vivo Assessment of Dopaminergic Neurotoxicity in Rodents methyllycaconitine: involvement of nicotinic receptors. Neuropharmacology 51:885–895. https://doi.org/10.1016/j.neuropharm. 2006.05.032 6. Chipana C, Torres I, Camarasa J et al (2008) Memantine protects against amphetamine derivatives-induced neurotoxic damage in rodents. Neuropharmacology 54:1254–1263. https://doi.org/10.1016/j.neuropharm. 2008.04.003 7. Chipana C, Garcı´a-Rate´s S, Camarasa J et al (2008) Different oxidative profile and nicotinic receptor interaction of amphetamine and 3,4-methylenedioxy-methamphetamine. Neurochem Int 52:401–410. https://doi.org/10. 1016/j.neuint.2007.07.016 8. Chipana C, Camarasa J, Pubill D, Escubedo E (2008) Memantine prevents MDMA-induced neurotoxicity. Neurotoxicology 29:179–183. https://doi.org/10.1016/j.neuro.2007.09. 005 9. Cadet JL, Krasnova IN (2009) Molecular bases of methamphetamine-induced neurodegeneration. Int Rev Neurobiol 88:101–119. https:// doi.org/10.1016/S0074-7742(09)88005-7 10. Song B-J, Moon K-H, V Upreti V et al (2010) Mechanisms of MDMA (ecstasy)-induced oxidative stress, mitochondrial dysfunction, and organ damage. Curr Pharm Biotechnol 11:434–443. https://doi.org/10.2174/ 138920110791591436 11. Yamamoto BK, Moszczynska A, Gudelsky GA (2010) Amphetamine toxicities: classical and emerging mechanisms. Ann N Y Acad Sci 1187:101–121 12. Marshall V, Grosset DG (2003) Role of dopamine transporter imaging in the diagnosis of atypical tremor disorders. Mov Disord 18: S22–S27. https://doi.org/10.1002/mds. 10574 13. Tissingh G, Bergmans P, Booij J et al (1998) Drug-naive patients with Parkinson’s disease in Hoehn and Yahr stages I and II show a bilateral decrease in striatal dopamine transporters as revealed by [123I]β-CIT SPECT. J Neurol 245:14–20. https://doi.org/10.1007/ s004150050168 14. Volkow ND, Chang L, Wang GJ et al (2001) Association of dopamine transporter reduction with psychomotor impairment in methamphetamine abusers. Am J Psychiatry 158:377–382. https://doi.org/10.1176/appi.ajp.158.3.377 15. Bannon MJ (2005) The dopamine transporter: role in neurotoxicity and human disease. Toxicol Appl Pharmacol 204:355–360. https:// doi.org/10.1016/j.taap.2004.08.013

457

16. Wong DF, Harris JC, Naidu S et al (1996) Dopamine transporters are markedly reduced in Lesch-Nyhan disease in vivo. Proc Natl Acad Sci U S A 93:5539–5543. https://doi.org/10. 1073/pnas.93.11.5539 17. Todd G, Noyes C, Flavel SC et al (2013) Illicit stimulant use is associated with abnormal substantia nigra morphology in humans. PLoS One 8:e56438. https://doi.org/10.1371/ journal.pone.0056438 18. Ellison G, Eison M, Huberman H, Daniel F (1978) Long-term changes in dopaminergic innervation of caudate nucleus after continuous amphetamine administration. Science (80) 201:276–278. https://doi.org/10.1126/sci ence.26975 19. Bu¨ttner A (2011) Review: the neuropathology of drug abuse. Neuropathol Appl Neurobiol 37:118–134 20. Hirano S (2012) Western blot analysis. Humana Press, Totowa, NJ, pp 87–97 21. Kurien BT, Hal Scofield R (2015) Western blotting: an introduction. In: Western blotting: methods and protocols. Springer, New York, pp 17–30 22. Taylor SC, Posch A (2014) The design of a quantitative western blot experiment. Biomed Res Int 2014 23. Lebel CP, Bondy SC (1990) Sensitive and rapid quantitation of oxygen reactive species formation in rat synaptosomes. Neurochem Int 17:435–440. https://doi.org/10.1016/ 0197-0186(90)90025-o 24. Mai HN, Sharma N, Shin E-J et al (2018) Exposure to far infrared ray protects methamphetamine-induced behavioral sensitization in glutathione peroxidase-1 knockout mice via attenuating mitochondrial burdens and dopamine D1 receptor activation. Neurochem Res 43:1118–1135. https://doi.org/10. 1007/s11064-018-2528-5 25. Coffey LL, Reith MEA (1994) [3H]WIN 35,428 binding to the dopamine uptake carrier. I. Effect of tonicity and buffer composition. J Neurosci Methods 51:23–30. https:// doi.org/10.1016/0165-0270(94)90022-1 26. Lo´pez-Arnau R, Martı´nez-Clemente J, Abad S et al (2014) Repeated doses of methylone, a new drug of abuse, induce changes in serotonin and dopamine systems in the mouse. Psychopharmacology 231:3119–3129. https://doi. org/10.1007/s00213-014-3493-6 27. Martı´nez-Clemente J, Lo´pez-Arnau R, Abad S et al (2014) Dose and time-dependent selective neurotoxicity induced by mephedrone in mice. PLoS One 9:e99002. https://doi.org/10. 1371/journal.pone.0099002

Chapter 21 Quantification of Oligodendrocytes and Myelin in Human iPSC-Derived 3D Brain Cell Cultures (BrainSpheres) David Pamies, Megan Chesnut, He´le`ne Paschoud, Marie-Gabrielle Zurich, Thomas Hartung, and Helena T. Hogberg Abstract Myelination is considered a critical process in the development of the vertebrate brain. This process is susceptible and can be affected by exposure to developmental neurotoxicants or by numerous diseases (e.g., Schizophrenia, bipolar disorder, amyotrophic lateral sclerosis). Studying human myelination has been very difficult due to the lack of in vitro human models capable of reproducing this process. A human 3D iPSCderived brain model (also called BrainSpheres—BS), developed by Johns Hopkins University, is a multicellular culture that includes different neuronal and glial cell types such as neurons, astrocytes, and oligodendrocytes, and is able to mimic human myelination in vitro. Here, we describe all the methods developed in this model in the last years to quantify oligodendrocytes and myelination. Application of Computer-assisted Evaluation of Myelin (CEM) (developed by Kertman et al.) and other immunochemistry quantification methods are here adapted to a 3D culture BrainSpheres. Key words Myelin, Oligodendrocytes, IPSC, BrainSpheres, Organotypic model, 3D in vitro culture

1

Introduction During the development of the nervous system, the processes of glial cells (oligodendrocytes in the central nervous system (CNS) and Schwann cells in the peripheral nervous system (PNS)) wrap membrane around axons to form the myelin sheaths. Myelin is a lipid-rich substance that is produced with the objective to isolate the axons to increase speed and efficiency of the electrical impulse in neurons and is essential for the function of the nervous system [1]. The myelin sheath is discontinuous, and the gaps between the segments of the sheath, called nodes of Ranvier, contain sodium channels [1–4]. Myelination allows fast saltatory conduction of action potentials along these nodes [2, 4] that increases the speed of an action potential along a myelinated axon approximately 10–100 times [5]. This increase reduces the need for ATP-dependent sodium-potassium exchange required for

Jordi Llorens and Marta Barenys (eds.), Experimental Neurotoxicology Methods, Neuromethods, vol. 172, https://doi.org/10.1007/978-1-0716-1637-6_21, © Springer Science+Business Media, LLC, part of Springer Nature 2021

459

460

David Pamies et al.

maintenance of the resting membrane potential along axons [6]. Myelin is mostly comprised of lipids (70%) and proteins (30%), and is formed by multiple protein-lipid layers [3, 7, 8]. The main structural proteins of myelin are myelin basic protein (MBP), proteolipid protein (PLP), 20 30 -cyclic-nucleotide 30 -phospodiesterase (CNP), myelin-associated glycoprotein (MAG), and myelin oligodendrocyte glycoprotein (MOG) [5, 7], and the main structural lipids of myelin are phospholipids, glycolipids, and cholesterol [5, 7–9]. Myelinating oligodendrocytes participate in neuronal signaling networks [10–12] and have a symbiotic relationship with the axon [13], promoting axonal growth, integrity, and survival [2, 6, 13–18]. In addition, a recent study has also demonstrated that oligodendrocytes play a role in inflammation [19]. Myelin is also metabolically active, maintains its own structure [20], regulates the ion composition and fluid volume around axons [21], and releases molecular factors that inhibit axon development and synapse formation, thus contributing to neural plasticity [22]. Areas in the CNS with mainly myelinated axons are termed white matter. Diffusion tensor imaging studies, which allow the mapping of white matter by tractography, have associated its maturation during childhood with improved language, memory [23], and reading performance [24]. Likewise, the failure to form or maintain myelin can disrupt neuronal signal transmission or trigger degradation of axons, and a reduction in the velocity of action potentials can lead to physical or mental disability [4]. Animal studies have implicated white matter deficits during critical periods in development in learning and memory impairments [25, 26] as well as an inability to learn motor skills [27]. Moreover, many human neurologic diseases have been associated with myelin deficits, white matter injury, or dysfunctional oligodendrocytes [4, 28]. These include schizophrenia [29–33], bipolar disorder [32], amyotrophic lateral sclerosis (ALS) [34, 35], depression [36], periventricular leukomalacia (PVL) [37, 38], which can result in cerebral palsy [39], and multiple sclerosis [40], among others. There are concerns about the increase of developmental diseases, such as autism and attention deficit hyperactivity disorder (ADHD), and their possible link to exposure to environmental chemicals [41–43]. However, only few developmental neurotoxicity (DNT) chemicals have been identified, as the assessment of environmental chemicals for DNT is not done routinely due to the high testing costs resulting from the existing guidelines [44, 45]. In addition, there are concerns regarding the relevance of these tests to human toxicity. Indeed, current DNT guidelines are based on animal testing where the dosage is performed in a different way than in the human exposure scenario. Moreover, the assessment is based on behavior and histology of the animals that are difficult to interpret and translate to human outcomes. In

Quantification of Oligodendrocytes and Myelin in Human iPSC-Derived 3D. . .

461

addition, the evaluation of DNT induced by exposure to environmental chemical depends not only on the duration and dose but also on the developmental stage of the brain at the time of exposure [46]. In order to better evaluate DNT, experts have suggested a battery of in vitro human test methods integrated into Integrated Approaches to Testing and Assessment (IATAs) to routinely evaluate compounds [47–49]. This battery must cover the key events (KE) of neurodevelopment at the cellular level (proliferation, migration, apoptosis, differentiation, synaptogenesis, neurite growth, myelination,neuronal network formation and function) [49]. European Food Safety Authority (EFSA) has published a detailed report on the evaluation of the currently available in vitro test methods suitable for DNT testing [50] and their readiness for regulatory application and the development of IATAs was recently discussed [51]. A first testing battery of in vitro and non-mammalian assays for DNT has been selected and is currently challenged with approximately 100 chemicals. At the same time, the Organisation for Economic Co-operation and Development (OECD) is in preparation of a guidance document that will describe the testing battery, its usage, and interpretation. However, one KE of neurodevelopment, the myelination process has been a challenge to study in vitro and is currently missing in the testing battery. Very few in vitro models have shown de novo myelination. A human 3D iPSC-derived brain model (also called BrainSpheres— BS), created at Johns Hopkins University, has shown to produce a multicellular culture that includes different neuronal and glial cell types, such as astrocytes and very importantly oligodendrocytes [52] mimicking the histology of the central nervous system. Several markers for oligodendrocytes (CNP, O1, O4, NOGOA, MOG, MBP) have been evidenced by immunohistochemistry and qRT-PCR [52, 53]. In BrainSpheres up to 40% of the axons are myelinated [52], with a relatively high level of compaction of the myelin sheath, as demonstrated by electron microscopy [52]. Electron microscopy demonstrated that the myelin sheaths are able to wrap axonal structures [52]. Here, we describe a method to assess myelination by quantifying MBP staining in human cells (Fig. 1). The protocol presented is adapted from a recent publication using mouse embryonic stem cells by the Cage group [25] and applied to BrainSpheres [52, 54].

2 2.1

Method iPSC Generation

iPSCs are cultured on irradiated mouse embryonic fibroblasts (MEFs) in human embryonic stem cell (hESC) medium comprising D-MEM/F12 (Invitrogen), 20% KnockOut™ Serum Replacement (KSR, Invitrogen), 2 mM L-glutamine (Invitrogen), 100μM MEM

462

David Pamies et al.

Fig. 1 Diagram of myelin quantification workflow. CEM computer-assisted evaluation of myelin formation

NEAA (Invitrogen), 100μM β-mercaptoethanol (Invitrogen), and 10 ng/mL human basic FGF (bFGF, PeproTech). iPSCs colonies are manually picked after 3–6 weeks of expansion for characterization. Only iPSCs with passage 20 are used for NPC differentiation. Medium is changed daily and iPSC lines are passaged using collagenase (Invitrogen, 1 mg/mL in D-MEM/F12 for 1 h at 37  C). Cell lines need to be karyotyped and mycoplasma-free. Feeder-free media can also be used. iPSCs have also been successfully cultured in feeder-free media, using mTeSR™ Plus ( STEM CELLS techonologies) on Matrigel (Corning Life Sciences) coated plates. 2.2

NPC Generation

BrainSpheres (BS) are formed from neural progenitor cells (NPCs). NPCs can be generated from pluripotent stem cells (PSC) using different protocols. In our hands, two protocols have been used to generate NPCs; however, other protocols could also be appropriate. NPCs are generated using previously published protocols based on embryonic bodies formation [55], or using the protocol “Induction of Neural Stem Cells from Human Pluripotent Stem Cells Using PSC Neural Induction Medium” (Gibco, Publication Number MAN0008031). Depending on which differentiation protocol is used, two different expansion NPC medias are applied. StemPro® NSC SFM (Life Technologies) is used for NPCs derived from EB and Neural Induction Expansion Medium is used for the NPCs derived from the Gibco protocol. Both protocols have shown to produce quality NPCs and subsequently BrainSpheres. NPCs can be cryopreserved. Specifications can be found in “Cryopreserve NSCs” section in the Gibco protocol.

Quantification of Oligodendrocytes and Myelin in Human iPSC-Derived 3D. . .

463

2.3 BrainSpheres Differentiation

Preparation of the organotypic human BrainSphere model follows a protocol that has been described by Pamies et al. (2017) [52]. Briefly, NPCs thawed from frozen stocks are expanded for 2–3 weeks, mechanically scraped, and dissociated into a single cell suspension. Accutase® can also be used to detach the NPCs and make the single cell solution. Two million cells per well (counted using TrypanBlue (GIBCO) and an automated cell counter (Countess; Invitrogen)) are plated in non-coated 6-well plates in 2 mL of NPC medium and incubated at 37  C, 5% CO2 as freefloating cultures under constant gyratory shaking (Kuhner or Thermo Scientific MaxQ 2000 CO2) at 88 rpm, 19 mm orbit. After 2 days, NPCs medium is exchanged to differentiation medium containing Neurobasal® electro medium (GIBCO) supplemented with 1% Penicillin Streptomycin Glutamine (GIBCO), 1% Glutamax (GIBCO), 1% B-27®Electrophysiology supplement (GIBCO), 0.02μg/mL human recombinant glial-cell derived neurotrophic factor (GDNF, Gemini), and 0.02μg/mL human recombinant brain-derived neurotrophic factor (BDNF, Gemini). Cultures are maintained for 8 weeks at 37  C, 5% CO2 under constant gyratory shaking at 88 rpm, 19 mm orbit with medium exchange 3 times per week [52]. The treatment of the compound of interest is from week 7 to week 8. No freezing protocol has been used successfully in BrainSpheres yet.

2.4 Chemical Exposure

Cuprizone (CAS number: 370-81-0), a well-known demyelinating compound [56, 57], was used as a reference compound to establish the method. Cuprizone-induced toxicity has been extensively used as a model for demyelinating diseases [56, 57]. Cuprizone oxalic acid bis(cyclohexylidene hydrazide) (SIGMA) produced oligodendrocytes cell death resulting in demyelination [58]. BS are exposed at 7 weeks of the differentiation process. Samples are collected at 8 weeks (after 1 week exposure).

2.5 Immunohistochemistry and Confocal Microscopy

Confocal microscopy is required to perform myelin quantification according to this method (Fig. 2). For immunohistochemistry, BS are exposed to test compounds from 7 to 8 weeks of differentiation with the compound solution renewed at each media exchanged (3 times per week). After the treatment, aggregates are collected with 1 mL micropipette and collected into 1.5 mL Eppendorf tubes. Media are removed and BS aggregates are washed two times with PBS. PBS used was always without calcium chloride and magnesium chloride (Gibco 14,190-144). Afterwards, BS are fixed for 1 h in 4% paraformaldehyde. After the fixation, BS are washed three times in PBS and incubated for 1 h in blocking solution (5% normal goat serum (NGS) in PBS with 0.4% Triton X-100). After the blocking step, BS are incubated at 4  C for 48 h with a combination of primary antibodies (NF200, MBP, PLPC1, OLIG1 or O4, Table 1) in 5% NGS, 0.1% Triton X-100 in PBS.

464

David Pamies et al.

Fig. 2 Myelin quantification on BrainSpheres by CEM. BrainSpheres have been differentiated for 7 weeks. Cuprizone was added to differentiation media to reach a final concentration of 50μM. BrainSpheres were treated for 7 days and collected for immunohistochemistry for (a) PLPC1 and NF200, (b) MBP and NF200, (c) total fluorescence quantification of PLPC1 staining, (d) CEM quantification of MBP staining, Ave average Table 1 Primary antibodies Antibody

Host

Type

Source

Code

Dilution

NF200

Rabbit

Polyclonal

Sigma

N4142

1:500

MBP/SMI99

Mouse

Monoclonal

BioLegend

808,401

1:200

PLPC1

Mouse

Monoclonal

BioRad

MCA839G

1:200

O4

Mouse

Monoclonal

R&D Systems

MAB1326

1:200

OLIG1

Mouse

Monoclonal

Sigma

MAB5540

1:200

Subsequently, BS are washed in PBS three times and incubated with secondary antibody for 1 h in PBS with 5% NGS at room temperature using the proper combination of secondary antibodies (e.g., goat anti-rabbit secondary antibody conjugated with Alexa 594 and goat anti-mouse secondary antibody conjugated with Alexa 488 (Molecular Probes)). For nuclei staining different dyes can be used, including DRAQ5 dye (Cell Signaling; 1:5000 in PBS), Hoechst 33342 (1:10,000 in PBS), or DAPI (1μg/mL, in PBS) for 1 h incubation. After nuclei staining, BS are washed two times in PBS and mounted on slides with coverslips and Prolong Gold antifade reagent (Molecular Probes). Images are taken using a Zeiss UV-LSM 510 confocal microscope or similar. At least

Quantification of Oligodendrocytes and Myelin in Human iPSC-Derived 3D. . .

465

10 pictures per condition are used for the imaging analysis, each photo taken in a different sphere (10 BrainSpheres).

3

Computer-Assisted Evaluation of Myelin (CEM) For myelin quantification, we use Computer-assisted Evaluation of Myelin (CEM). CEM is an ImageJ plugin developed by Kerman et al., 2015 [59] that can be downloaded for free.1 In the supplementary material of the Kerman et al., 2015 article a tutorial demonstrates how to use the plugin. In addition, we developed a macro to automatically generate the binary pictures necessary for the analysis (Table 2). Briefly, create a file directory and transfer all images (LSM files) you would like to analyze to one folder. Create a new folder for the binary single channel images that will be generated in this protocol. It is important that neither of these folders is contained within the other. In order to determine the optimal threshold settings for your images, open an image of a control sample and an image from a treated sample (positive control, e.g., Cuprizone) where less myelination is observable. Open the Brightness/Contrast window in ImageJ. Adjust the threshold of the green and red channels for both images by changing the minimum threshold value in the Brightness/Contrast window. Check that background is reduced, but that important features are still observable. Record which values you have selected for the red and green channels. Next, determine which channel numbers correspond to your markers. Record these channel numbers. In the example below, MBP (green) is channel 1, NF (red) is channel 2, and Hoechst 33342 (blue) is channel 3. Edit the macro text in Box 1 to reflect the correct channel numbers (highlighted in the color of the channel, Hoechst, MBP, NF) and replace the words “min” and “max” (highlighted in yellow) with the minimum and maximum threshold values chosen for the green and red channels. If the images from the experiment do not contain a channel with nuclear staining, delete the text for this portion of the macro (red font). Afterward, run the batch binary conversion macro and follow. It is also possible to use total fluorescence for other antibodies related to myelin different than MBP such as PLPC1 (Fig. 3a). After having obtained the files with the binary pictures, follow CEM instructions for myelin quantification. In Fig. 2, we see an example of demyelination after cuprizone treatment for 1 week, having been measured by total fluorescence (Fig. 3a) or CEM (Fig. 3b).

1

http://www.biologists.com/DEV_Movies/DEV116517/DEV116517_Appendix%20S2-CEM%20package. zip

Table 2 Macro for ImageJ to generate binary pictures for CEM macro "Batch binary macro for myelin quantification by MC" { // chose the folders where the photos are stored and where the binary folders will be stored // dir1 = getDirectory("Choose Source Directory"); dir2 = getDirectory("Choose Destination Directory"); list = getFileList(dir1); // imageJ batch mode // setBatchMode(true); for (i=0; i